Professional Documents
Culture Documents
Contents i
by
Peter Watermeyer
© P Watermeyer
A CIP catalogue record for this book is available from the British Library.
The publishers are not responsible for any statement made in this
publication. Data, discussion, and conclusions developed by the Author
are for information only and are not intended for use without
independent substantiating investigation on the part of the potential
users. Opinions expressed are those of the Author and are not necessarily
those of the Institution of Mechanical Engineers or its publishers.
Foreword ix
Acknowledgements x
Chapter 1 Introduction 1
1.1 What’s so special about process plant projects? 1
1.2 The structure and components of this book 3
1.3 Methodology of presentation 5
1.4 Getting it right 7
First Cycle – A Process Plant and a Project
Chapter 2 A Process Plant 13
2.1 Basic process design elements 13
2.2 The processed materials and the process 18
2.3 The process design/detail design interface 20
Chapter 3 A Project and its Management: A Brief Overview 23
3.1 The project 23
3.2 Starting the project 25
3.3 Managing the project 27
Chapter 4 The Engineering Work and Its Management 31
4.1 Planning the engineering work 31
4.2 Doing the engineering work 40
4.3 Managing schedule, cost, and quality 41
4.4 Co-ordinating engineering work 44
4.5 Management of engineering resources 46
Second Cycle – Environment
Chapter 5 The Project’s Industrial Environment 51
5.1 The industry and the client 51
5.2 The client and the project management 55
5.3 The process plant project industry:
the ‘indirect cost’ of a plant 56
Chapter 6 The Commercial Environment 59
6.1 Principles of procurement and contract 59
Chapter 7 The Contracting Environment 65
7.1 Ways of building a plant 65
7.2 The engineering contractor 70
7.3 The project engineer 72
7.4 Conclusion 75
iv Contents
Chapter 1
Introduction
This book is intended to assist people who design and build process
plant, and people who participate in multi-disciplinary engineering
projects in general. The book is aimed in particular at the project engineer,
or team of lead engineers, who inevitably hold such projects together,
being at the heart of the information generation system which shapes
and guides the project.
The design and construction of process plant covers an exceedingly
wide field of performance and knowledge. Considering engineering
topics alone, those which impact process plant design include the work
of very many specialist technical branches. But this is only part of what
has to be considered in successful project engineering. Interwoven with
the engineering and design of the plant are many commercial, construc-
tional, financial, and social considerations. It may be satisfying for the
technical purist to focus exclusively on engineering issues, but the
engineer who does so is likely either to be limited to a subordinate role,
or to be part of an unsuccessful enterprise – engineering considered on
its own becomes an academic pastime.
The content of the book is therefore intended to address that mixed
bag of technical, commercial, managerial, and behavioural issues which
constitute the actual job content of the practising engineer, and are
peculiar to this industry.
case for plants which are very similar to existing units. Within the
following text may be found some helpful aspects of technical advice
for such projects, but the work must not be over-complicated. Any
higher-level issues of strategy and management are best addressed
(if at all) at the level of organization which encompasses the small
project, for example the corporation which requires the project, or a
contracting organization which delivers such projects. The following
text is primarily addressed at relatively large and complex projects,
requiring the interaction of a number of skills – technical, commercial,
and constructional.
Compared to the general field of engineering design and construction,
the main differentiating factors which have to be addressed in these
projects are the following.
• The unique design of each plant is the inevitable consequence of the
need to optimize each application to its unique circumstances of
feedstock, product, capacity, and environment.
• Plants are built around hundreds of items of proprietary processing
equipment. The plant design must interface exactly with the
operational characteristics and dimensions of these individual items.
The interface information can only be finalized when the commercial
agreements with equipment suppliers have been concluded.1
• Both the plant design and its construction employ many types of
specialist, who must interact at thousands of interfaces (and often
work simultaneously in the same plant space envelope) to produce a
co-ordinated product.
• Plant operation can be hazardous. The elimination or control of
hazards, and the establishment of safe operating practices, are
prerequisites of plant operation, requiring priority attention during
plant design. Environmental impact is also invariably an issue.
• Technology development is rapid and continuous. This greatly
impacts on the uniqueness of process and plant design (the first item
above), but also often impacts on project execution – there is often a
desire to accommodate changes while design and construction are in
progress.
1
This is at least the common practice in free-market economies. In communist countries,
it was (and still is) often the practice to ‘centralize’ equipment designs, so that the plant
designer can simply choose a completely detail-designed equipment item from a
catalogue. This simplifies the plant designer’s job,but often at great cost at the construction
stage, when there is no commitment from a supplier to match the project programme or
budget. Also, it shifts responsibility for ‘fitness for purpose’ of the item.
Introduction 3
At the level above, level two, reside the technical focus and project
control, which give direction to all the detailed work. The centrepiece
of this is the process technology package. The technological content
of such packages is outside the scope of this book – it is the domain
of many experts and organizations, embracing diverse fields of technical
specialization such as catalytic conversion and hydro-metallurgy, and
niches within these fields which yet consume entire careers. We will
rather (and briefly) address the generalized make-up of such packages,
of how they relate to the plant to be built and to the project around
them. In addition, we will discuss the system of engineering and infor-
mation management (also at level two), which governs the performance
of level one activities.
Not particularly addressed at level two, because of the focus of this
book, are the methodologies of managing procurement, logistics,
contracts, finance, and construction. The reader with greater project
management aspirations is advised to refer to specialized texts on these
subjects. We will, however, discuss the principal interfaces with
engineering work and its management.
At level three, we have a management system for the entire project.
This is conventionally broken down into three or four components,
namely management of scope, quality, cost, and schedule. (Scope and
quality may be regarded as a single issue, a practice not recommended
by the author.) Health and safety considerations may properly also be
managed at this overview level, and must be included in any such text.
At level four, there is only one item, project strategy. This has to
ensure that the project is correctly conceived (technically, commercially,
economically, socially) and embodies the skeleton of the over-arching
plan which will ensure that the goods are delivered in the optimum
fashion. Inevitably the strategy must deal with the issues of relationship
management between the principal stakeholders, for example the plant
owner (usually, a complexity of people and interests) and contractors,
including possibly a single lump sum or managing contractor, and a
variety of sub-contractors and suppliers. The strategies are bound to be
different for each stakeholder, reflecting the basic question of ‘What do
I want out of this project?’
There is a significant body of professional opinion that such con-
siderations of strategy are not appropriate in the context of what should
be the technical field of engineering. (In the author’s experience, there
are also several senior executives who are relatively ignorant of the links
between engineering and strategy, and feel quite threatened!) There are
those who argue that management of large projects, whether or not
Introduction 5
2
Throughout the text, the male gender is intended to include the female. This is done
purely for the purposes of simplification.
8 Handbook for Process Plant Project Engineers
First Cycle
A Process Plant and a Project
This page intentionally left blank
A Process Plant 13
Chapter 2
A Process Plant
The process flowsheets represent the process rather than the details of
the plant. The latter are shown in ‘P&I’1 diagrams, which depict all
1
Pipeline and instrumentation, although sometimes described as process and instrument-
ation; but P&I has become an accepted international multilingual expression. Some
engineers use ‘mechanical flow diagrams’, which do not show much instrumentation,
and ‘control and instrumentation diagrams’, which focus as the name implies, and no
doubt such presentation is appropriate for certain applications; but P&IDs usually
suffice.
16 Handbook for Process Plant Project Engineers
plant equipment items, including their drive motors, all pipelines and
valves (including their sizes), and all instruments and control loops.
Utility flowsheets and diagrams are often presented separately.
Plants may operate by batch production, in which the plant processes
a quantum of feed per cycle, and stops at the end of each cycle for
removal of the product and replacement of the feed. Alternatively,
plants may operate continuously, 24 h per day, without stopping; and
there are hybrid plants, or hybrid unit operations within plants, which
are described as semi-continuous, in that the internal operation is cyclical
but the cycles follow continuously, one after the other, with little operator
intervention.
The critical performance factors for a process plant – the factors
which determine its fitness for purpose and its effectiveness (and against
which its designers’ performance is measured) – include the following.
• Feedstock transformation as specified. Product characteristics
should be within a specified range corresponding to feedstock
characteristics within a specified range, and capacity (throughput)
should be within the range required for feed and for product. From
the feed and product capacity may be derived the recovery, or yield
of product per unit feed. Alternatively, the recovery and input or
output may be stated, and the output or input respectively may be
derived.2
• Cost of production, often expressed per unit of feed or product. The
cost components include capital amortization and interest, plant
operators’ salaries, maintenance materials and labour, purchased
utilities, process reagents, insurance, etc. There may also be fees
payable to process technology licensors. The capital cost component
is often quoted separately as a stand-alone criterion.
• Plant reliability and availability. Reliability is the predictability of
plant operation as planned, whereas availability is the proportion
of time for which the plant is in a condition whereby operation
(to acceptable standards) is possible. Availability may be less
than 100 per cent because of planned outages for maintenance,
for example one 3-week shutdown per 2-year cycle, or because of
shutdowns caused by lack of reliability, or (invariably to some
degree) both.
• Safety of construction and operation. This is assessed at the design
stage by formalized hazard analysis for the process and by hazard
2
For some plants, there may also be a specification linking the feedstock or product
specification (or grade) to the capacity and/or the recovery.
A Process Plant
17
Fig. 2.2 P&I diagram
18 Handbook for Process Plant Project Engineers
appreciated that the engineering of the plant can be separated into two
parts. The first part is the process technology which would be applicable
to a plant built on any site,3 and in general utilizing any combination
of appropriate equipment vendors. The second part is the complete
engineering design of the plant, incorporating actual proprietary
equipment designs, locally available construction materials, local design
practices and regulations, customized design features required by the
particular client and his operation and maintenance staff, and layout
and other design features necessary for the plant site. The second part is
often referred to as the ‘detailed engineering’. There is an area of potential
overlap between the two; in particular, the process package can be
expanded to a ‘basic engineering’ package, which includes the essence
of detailed engineering (such as well-developed plant layouts and
equipment lists) as well as the process package.
It needs to be understood that it is difficult to include all the required
knowledge of a particular type of process into a stand-alone package. It
is even more difficult in the case of complex processes or newly devel-
oped processes. In practice, when the process technology provider (or
licensor) is separate from the detailed engineering organization, it is
necessary to have most of the important design details reviewed and
approved by the technology provider, to ensure that the process require-
ments have been correctly interpreted. Further process technology input
is also needed in the preparation of detailed plant operation manuals
and plant commissioning, and sometimes into the plant construction.
3
This is an over-simplification. In practice, the process design package usually has to be
customized to take into account local factors which affect the process, such as ambient
conditions, feedstock and reagent variations, and properties of available utilities.
This page intentionally left blank
A Project and its Management: A Brief Overview 23
Chapter 3
project is initiated with parameters set by the study. Because our focus is
on the design and construction of the plant, we shall give minimal attention
here to the work done prior to the decision to go ahead. These aspects
are given more detailed review in Chapter 9, Studies and Proposals.
A project may be a major new enterprise such as an oil refinery, or a
small modification to an existing plant. For the latter case, it is evident
that the desired objective can be achieved quickly and informally
without elaborate procedures. As the size and complexity increase, the
need for a more formal approach becomes apparent, for many reasons.
More people are involved, more interacting components have to be
co-ordinated, and the investors demand more detailed reporting. The
size of project at which more formalized procedures become necessary
depends very much on the ability and skills of the project manager and
the demands of the client. In general we will be addressing the needs of
the large project, on the basis that engineers who are experienced in the
bigger picture will understand what shortcuts and simplifications are
reasonable on smaller projects.
Pre-project work starts with an idea or concept which the client has
decided to develop. The concept and design of the final process plant
progresses in cycles of increasing definition. Initially a study is made, in
which the concept is technically developed, optimized, and analysed as a
business proposition; the analysis includes considerations of technical
and commercial risk, capital and operational cost, product value, and
return on investment. A report is prepared; if the conclusions are
acceptable to the client, he may authorize the implementation of the
project. Alternatively, he may authorize more funds for further conceptual
development, or, of course, abandon the concept. Authorization of the
implementation of the project invariably implies the expectation of a
plant which will perform within specified limits, and be built in accordance
with certain standards, within a promised budget and schedule.
There is clearly an amount of pre-project engineering work necessary
to achieve the required degree of technical definition, costing, and
schedule analysis required for authorization. Prior to the decision to
implement the project, there is a natural reluctance to spend any more
funds than are absolutely necessary to complete the feasibility study, as
there may be no project. This reluctance is tempered by the need for
accuracy ⫺ evidently, the further the engineering of the plant is developed,
the greater the confidence in the accuracy of the study report. In addition,
if the client is confident that the study will lead to a project without much
further conceptual design development, he may be willing to commit
more funds prior to final authorization, in order to expedite the project.
A Project and its Management: A Brief Overview 25
1
To be more accurate, it is not possible for the type of project work envisaged here,
where a defined amount of work is required to be performed for a pre-determined cost.
It is of course possible to carry out elastic-scope and undefined cost or schedule work
without much of a baseline.
26 Handbook for Process Plant Project Engineers
Chapter 4
• the performance of the plant (what must the plant produce, and from
what input resources, when, and how?);
• the processes by which the plant will function;
• all the special or local circumstances which will affect the construction
or function of the plant; and
• any other specific requirements of the client.
To make sure that this ‘clear understanding’ is correct, is comprehensive,
and describes exactly ‘what the customer wants’, the obvious procedure
is to write it down as concisely as possible, review it thoroughly with the
plant owner and operator, and get it formally approved before proceeding
with design work.
In addition to the essential requirements mentioned in the previous
paragraph, the engineer is confronted with choices of methodology
during the design development, for instance which design codes to use,
if this has not been specified by the client. The potential impact of
choice of methods and codes must be considered. Often, there is little to
differentiate between the possibilities other than the engineer’s familiarity,
but a choice must be made: a design based on a hotch-potch of codes is
a recipe for disaster.
Because the procedure of developing this ‘clear understanding’ and
choice of design methodology determines the entire design and con-
struction of the plant, it is obviously of prime importance and deserves
corresponding priority and effort. The performance and constructional
standards which define the plant with ‘clear understanding’, and the
design methods to be employed, will be termed the ‘plant (or project)
design criteria’. The documents and models subsequently produced to
facilitate purchase and construction (in accordance with the plant design
criteria) will be termed the ‘design documentation’.
For the development of the design criteria, two scenarios need to be
considered (there are many hybrids in practice). Firstly, the client may
not know in detail what he wants, or he may know what is required but
not have fully thought it out and written it down. For this first case it is
the project engineer’s job to establish the design criteria in conjunction
with the client, generally by a process of reviewing the choices available
and making recommendations, and to get the client’s approval. If he
does not do a thorough job of this, he risks at least wasted effort and
delay in the design process, when fully developed designs are rejected on
conceptual grounds, or at worst a final plant which is not what the
client wanted.
In the second scenario, the client defines what he wants in a detailed
The Engineering Work and Its Management 33
1
A ‘document’ is a piece of recorded in formation of any type (hard copy, electronic,
physical model). This subject is further developed in Chapter 20, Traditional
Documentation Control.
The Engineering Work and Its Management 35
2
This subject is discussed in Chapter 25, Communication.
36 Handbook for Process Plant Project Engineers
3
The making and control of assumptions, in order to expedite engineering, is one of the
key issues of project engineering in a competitive environment. This is discussed in
Chapter 27, Fast-Track Projects.
The Engineering Work and Its Management 41
4
This is developed further in Chapter 25, Communication.
The Engineering Work and Its Management 45
Careful review of the initial plan by the design team leaders shows up
where special steps need to be taken to reduce iterations or information
delays, and to modify the plan accordingly. Such steps typically include
the following.
• Adoption of conservative design parameters, which will not usually
be changed if subsequently proved to be over-conservative. Exam-
ples are: allow greater clearances within a structure, allow for
higher structural loads, and allow for higher driven machine power
requirements. Clearly the benefits of such actions have to be balanced
against the cost implications.
• More conceptual work to eliminate uncertainties, which may
otherwise result in design reiteration.
• Changing the design to a more standard or better-proven design.
This may include the purchase of a proprietary item or package, in
place of a customized project design.
• Looking for previous similar project experience, or a suitably
experienced consultant.
• Where uncertainties arise out of future procurement decisions,
changing the procurement policy. For example, by negotiation with
a selected supplier rather than competitive purchase, or, less radically,
by specification of design features which otherwise would be
determined by competition. The cost and commercial implications
obviously have to be evaluated.
During implementation of the engineering work, a similar range of
remedial steps may be considered if it becomes necessary to bring the
work back on schedule.
The greatest challenge to co-ordination usually lies in the drawing office.
The challenge is best illustrated by some of the problems to be avoided.
• Drawings chronically late for construction needs while design
problems are resolved.
• Support structures which do not fit the supported equipment.
• Clashes between piping and steelwork.
• Items which cannot be erected on site without being cut into pieces.
• Process equipment at the wrong elevation for flow of process materials.
• Inadequate plant access and maintainability.
In order to identify the actions needed to avoid such problems, we will
consider the causes.
• The first and probably most frequent cause lies in the drafting
process, or process by which the plant is modelled. The possible
46 Handbook for Process Plant Project Engineers
tools such as computers, and the facilities needed. From the detailed
work breakdown, which includes the resource type and hours needed for
each activity (or document), and from the required work completion
schedule, it is a simple matter to plot out the input hours required of
each resource category against time. There will inevitably be periods of
unacceptably high or low resource utilization, which are incompatible
with reasonable standards of work continuity, team-building, and resource
availability. The schedule usually has to be ‘smoothed’ by a process of
compromise to yield a schedule which can be resourced with reasonable
confidence and is considered to be optimal. The process of compromise
is made more difficult and challenging by the fact that the schedule
sequence logic is also subject to compromise if necessary, as discussed
previously, and in fact both aspects must be considered together.
The management of process plant engineering resource costs and
schedule durations is frequently problematical and occasionally
disastrous. Often the problems stem from a technical error, or series of
errors, which are discovered too late and require remedial action.
Equally often, however, the problems stem from management failure,
that is, from failure to plan the work and control its execution according
to the plan. The plan may be unachievable because the project team is
not sufficiently competent to meet the challenge, or because irresistible
(or insufficiently resisted) external factors dictated an over-optimistic
commitment.
Really disastrous overruns may occur if not only the work performance
but also the monitoring of progress against the plan is faulty, so that the
project manager has insufficient time to take appropriate measures. This
can be caused by simply leaving items of work off the plan; usually the
quality-related items such as layout reviews and corresponding re-work
allowances. Flawed designs go undetected, and remedial work is
deferred until the consequences of omission come home to roost.
Another frequent cause of over-optimistic progress reporting is failure
to measure accurately what work has been completed and what is
outstanding, because documents are regarded as complete when they are
subject to change (the consequences of which are routinely underesti-
mated). The experienced project engineer learns to ensure personally
that the plan and its monitoring retain their integrity in all these
regards, because a third-party planner (or scheduler), however well
trained, may have insufficient ‘feel’ for these and similar aspects.
This page intentionally left blank
The Project’s Industrial Environment 49
Second Cycle
Environment
This page intentionally left blank
The Project’s Industrial Environment 51
Chapter 5
other than sewage or minor detritus in slurry form (the most transport-
able slurry of all). The main emphasis when building such plants is on
reducing capital costs to a minimum, while remaining (just) within
whatever standards of product quality and construction are specified.
Innovation by technology advance (however minor), by structural
optimization, or by contracting method, is at a premium, with minimal
down-side potential (but some contractors still lose their shirts!).
Power plants, for the generation of electricity, are not normally
included under the definition of process plant. Certainly, when the
prime mover is a diesel engine, water turbine, or simple-cycle gas
turbine, the plant’s characteristics more closely resemble those of a
single piece of equipment than those of a process unit. However, a
major coal-fired power station, and the project to build it, have much in
common with process plant work: the system is built around a flowsheet
with many participating items of equipment and connecting pipework
and materials handling devices. In this case, the most important
differentiating characteristic is that the product – the kWh – is of
relatively low individual value, is required in massive and predictable
quantities over the future life of the plant, and cannot economically be
stored. As a consequence, there is a major emphasis on overall energy
efficiency and the factors which contribute towards it, and usually a
willingness to pay more in capital to achieve this goal. Capital is
relatively cheaply available for the perceived low risks involved. These
factors also contribute to a relative willingness to accept longer project
completion schedules, provided that some cost reduction is thereby
obtained, and also assuming that generation capacity growth has kept
up with the demand.
The comments above are of course highly superficial in relation to
industries which have developed technologies that people spend their
careers in acquiring. Our purpose here is not to discuss technology, but
its project application, and the main point to be made is that each type
of plant industry has developed in a way that has been driven by
characteristics related to the nature of the process and processed materials
and the scale of operation. The consequences of this development are
not restricted to technology, which may be presented in a package, but
extend to the attitudes and general behaviour of the people involved in
the client industry, in particular, their attitudes to innovation, work
standards, and cost reduction. These attitudes need to be understood for
successful project work. It also needs to be understood that there are
many anomalies; for example, petroleum companies that have acquired
mining interests may attempt to organize metallurgical plant projects
The Project’s Industrial Environment 55
or whether to save the associated capital cost increase and rely on external
contract maintenance. In other words, shift the cost of this service from
capital cost to operating cost. Similar scope variations are possible with
many nice-to-have items, such as standby equipment or maintenance-
saving features. If these practices are permitted, the cost estimate and
cost budget for the plant become simply self-fulfilling prophesies, and
the pressures of budget compliance are not nearly as great as in a
turnkey environment.
Aspects of such management practices are likely to appear in all
vertical relationships in the project hierarchy, from the chief executive
of the corporation that owns the plant down through the project team
and sub-contractors. A responsible party tends to hold those reporting
or sub-contracting to him accountable in detail, while attempting to
preserve the maximum of flexibility of performance target for himself.2
In the following pages we will in general assume that there is a
competitive environment, and that projects have to be completed in
accordance with technical and scope specifications, on time and within
budget. Those project practitioners who operate in less demanding
circumstances should quietly enjoy their good fortune.
the materials purchased and the labour hired for the job, and indirects
such as the costs of his own detailed engineering for the job and his own
contract management and overheads. In the case of a piece of equip-
ment such as a pump, the supplier’s costs may be broken down into the
directs of labour, material, and component costs on the one hand, and
the indirects of customized engineering, sales, factory overheads, and
order management costs on the other. The ‘direct’ costs can be further
broken down by investigating sub-contractors and component suppliers,
ad infinitum!
By carrying out such further breakdown, it is easy to reduce the aver-
age plant project direct cost proportion to below 50 per cent, with scope
for even further reduction for those interested in futile intellectual
exercises.
Defining what may be classed as an indirect cost is not just a question
of terminology. The actual work performance of many indirect cost
items may be moved between the project engineering and management
team and the fabrication and construction contractors, depending on
how the contractors’ workscopes are defined. In particular, this can be
done by purchasing either large, all-inclusive packages, or many small
items.
More cost-effective plant design may make certain pieces of equipment
redundant, leading to reduced direct costs as a result of improved
performance in the indirect cost sector. The reverse process, of performing
substandard engineering and management work and causing an increase
of direct costs, is also a frequent occurrence!
For many clients, and indeed for many project practitioners, the
problems of distinction between indirect and direct costs discussed
above are academic and irrelevant. To these people, the understanding
of what work is to be done for a given project, and how it is to be done,
is clear and unvarying. Any questioning on how the work can be broken
down between direct and indirect, and what constitutes an appropriate
proportion of indirect costs, is likely to be taken as an attempt to
disguise inefficiency. (‘We know that 16 per cent indirects is the right
figure for this type of plant.’) The issues surrounding indirect cost ratio
will be further debated at some length in the later pages, as they have an
important influence on how work is done. We will hope to convince the
reader that the reality is not as simple as the 16-per-centers would
have it, and that those who are prepared to acknowledge and study
the reality have scope for better performance than their more myopic
competitors.
The Commercial Environment 59
Chapter 6
1
This adjective is intended to imply that sham partnership arrangements are common in
practice.
The Commercial Environment 61
2
Originally, the text here was ‘... and pocket the difference.’ This has been amended, on
review, to clarify that the author is not advocating the practice (common in certain
environments) summarized by ‘5 per cent into my offshore bank account, and the order
is yours’.
The Commercial Environment 63
3
Possibly over cocktails ... and possibly while our young engineer is being told that he’s
‘the only real engineer in the office’. Manipulation is a two-way process!
This page intentionally left blank
The Commercial Environment 65
Chapter 7
purchase, which can be under client surveillance and direction, and only
the EPCM services costs need further attention as to how they will be
competitively obtained.
These services can be (and are) structured and paid for in a number of
ways:
• by the ‘manhour’ (still usually described as such at the time of writing,
but surely under threat of conversion to ‘personhour’?);
• by lump sum;
• by percentage of the constructed value of the plant;
• with bonus for good performance, and penalty for poor performance.
There are many variations; we will consider only the major factors
which influence the choice.
The performance of the EPCM services impacts directly on the DFC,
which is much greater. It is not in the client’s interests to save a relatively
small amount on the EPCM cost, and lose a larger amount on DFC
and plant performance as a result. This is the principal argument for
contracting on a manhour basis: there is no disincentive to the EPCM
contractor to spend more manhours when justified. Correspondingly
there is no incentive for efficiency, and while various ingenious formulae
have been drawn up to give such an incentive, it is difficult to overcome
the fundamental antithesis between restricting EPCM costs on the one
hand, and ensuring adequate work to get best value out of direct field
costs on the other. Another problem is that it is very difficult for the
client to compare contractors’ bids. In an hour’s worth of ‘manhour’,
how much is included of dedication, of efficient methodology, of truly
relevant experience, of real value?
An EPCM contract on a lump sum basis does not solve the problem
of antithesis – there is now every incentive for the contractor to mini-
mize the EPCM input, but none to promote better EPCM performance.
This goes even further in the contract, where the EPCM contractor is
reimbursed by a percentage of DFC; there is now a positive incentive to
raise the DFC. If this can be done in ways which reduce the EPCM
input, for instance by purchasing large packages or choosing more
expensive suppliers and contractors who require less surveillance or
expediting, the EPCM contractor scores twice.
One of the most common systems to seek a balance between the
conflicting needs is to split the EPCM work into two stages: the feasibility
study and the project. For the study work, the contractor is primarily
motivated by the need to secure the project work, which will follow the
study. Following the execution of the study, the project work should be
70 Handbook for Process Plant Project Engineers
1
Colloquially described as ‘bullshit’ in the industrial environment!
The Commercial Environment 73
7.4 Conclusion
Some of the views, concerns, and practices which have been discussed
may be regarded as confrontational and unnecessarily pessimistic of
human nature. In many situations this may be the case: engineering
work may be conducted in a harmonious and professional manner
without any client/contractor conflict; relationships of trust may be
developed without any abuse or loss of competitive performance.
However, the project engineer must be aware of the full range of
working relationships and their consequences, if only to reinforce his
determination to maintain the status quo of an existing relationship.
On the other hand, it is entirely possible for an engineer to find
himself in a situation which cannot be managed in terms of the logical
methodology discussed. For instance, to be locked into a contract with a
defined price, defined project schedule, undefined scope of work, and an
unreasonable client. The only response to this situation is firstly to
analyse and expose the basic unmanageability of the task, and then to
negotiate, preferably before doing too much work.
This page intentionally left blank
The Economic Environment 77
Chapter 8
In this chapter, we will further address and quantify some of the ‘project
environment’ issues raised in preceding chapters. They essentially relate
to two aspects of project work, firstly plant capital and operating costs,
and secondly the costs of project engineering and management work.
1
The interest can be calculated in different ways which are equally valid, provided
that the users of the calculated values understand the significance of the different calcu-
lation methods. One way is to use the interest corresponding to the cost of debt capital
to the client organization. Another way is to use the rate of return expected for equity
shareholders; another is to use the opportunity rate, the rate of return which may be
expected from alternative investment possibilities.
2
‘Direct’ in this context means all costs other than those arising from the capital cost
and associated interest.
78 Handbook for Process Plant Project Engineers
operating revenue to yield the net annual profit expected from the plant
venture.
Another method of analysis is to compute the net present value
(NPV) of the future plant operating margins (that is, product revenue
less direct operating cost), by discounting future earnings at an appro-
priate interest rate,3 and comparing the NPV with the estimated capital
cost. It is a simple mathematical exercise to demonstrate that the two
approaches are merely different presentations of the same information.
In the following, we will use the NPV approach, because it is easier to
utilize when making decisions on different plant features which have
capital/operating cost implications.
3
The same considerations about the rate apply.
4
Or increased revenue arising from more, or improved, product. If revenue is regarded
as a negative cost, the treatment is identical.
The Economic Environment 79
there is no need for further discussion. But typically at the study stage, it
will be the engineers’ responsibility to determine the correct balance.
Theoretically, this may be established by calculating the net present
value of a future-cost-saving option, allowing for the fact that the future
savings have to be discounted by a rate corresponding to the effective
interest rate, which is the cost to the investor of financing the option.
For large sums, arriving at the effective rate and making the decision is
often no simple matter; it may depend on the availability of venture
capital and, indeed, the additional finance may not be available whatever
the return. For smaller sums, it is usually possible to obtain a suitable
rate for the purposes of decision-making, and it is a simple matter to
compute the net present value according to the discounted returns over
the life of the plant.
For everyday decision-making, it is convenient to express the result in
terms of a time pay-back period by which to ratio annual cost-savings
to arrive at the discounted total of future savings. Having evaluated this
figure once for the project, it becomes a criterion which can quickly be
applied to each application.
As an example (there are variations according to the financial
presentation required), if
P = net present value of total future savings
S = net cost-saving per year
n = plant life (years)
i % = annual ‘effective interest rate’ (or discount rate)
Then if we define d by
100 - i
d=
100
P = S (1 + d + d 2 + d 3 + + d n -1 )
P - Pd = S (1 - d n )
Therefore
1 - dn
P=S
1- d
80 Handbook for Process Plant Project Engineers
P 1 - dn
= years
S 1- d
For instance, if the annual discount rate is 20 per cent (d = 0.8) and n is
15 years
P 1 - 0.815
= = 4.82 years
S 1 - 0.8
In the above, we have assumed that the benefits of ‘the feature’ are
available over the life of the plant, which implies that any additional
maintenance or periodic replacement costs to keep the feature effective
have been deducted in calculating the annual cost savings. Alternatively,
of course, if the proposed feature has a life of say 4 years, then one can
simply put n equal to 4 to evaluate the proposal. It is also a simple
matter to revise the algebra and series summation above to reflect a
different incidence of cash flow, for example reflecting commencement
of repayment in the second year after the capital outlay
d - dn
P=S
1- d
In the general case, if cash flow arising from the investment expenditure
commences a years after paying for it, (a may be zero or one as above,
and need not be an integer)
P da - dn
=
S 1- d
Taxation rules that affect plant finance are many. The most important
factor is usually the treatment of depreciation, which is frequently used
as an investment incentive by allowing the capital cost to be written off
against operating costs over an unrealistically short period, for example
3 years (notwithstanding that maintenance policies are usually to
maintain or enhance plant value in the first years of operation). The
effect of this is to make capital cheaper, by an amount which depends
on the discounted value of the tax saving.5 On the other hand, there
may be a tax ‘holiday’ over the period of initial operation, effectively
nullifying the benefits of depreciation. The taxation treatment of items
of operating expense and revenue may likewise have an effect on the
optimum plant configuration and design.
Low-interest finance packages, such as export incentives, may also
have a major effect on the cost of capital, and influence the design more
directly by restricting the sources of supply. Thus in the real world,
it is usually necessary to work out each year’s cash flow in detail, and
to apply actual taxation and item-specific capital interest, to reach the
correct trade-off for individual item design optimization. Failure to
ascertain the real impact of the fiscal and investment environment can
lead to quite uncompetitive designs.
There are some factors that affect the capital/operating cost dynamics
which are best left to the plant owner. One is the assessment of uninsur-
able risk, related to the possible effects of changes of cost inflation and
changes of the market for the plant’s product. The second is that in the
wider field of the client’s operations, especially for global companies, it
may be expeditious to incur apparently sub-optimal costs and profits at
isolated parts of the overall operation, in order to centralize profits in
one place rather than another, and thereby make significant tax gains.
However, if the correct attention and consultation is given at the
outset of the project or study, it is usually possible to arrive at a simple
set of rules, expressed in pay-off time, that makes financial analysis
relatively simple. (The reservations expressed above on the credibility of
certain technical aspects of the analysis remain.)
5
There is occasionally a sting in the tail here; if the plant performs subnormally in its
initial years, there may be no profits on which to reduce the tax.
84 Handbook for Process Plant Project Engineers
outside consultants, and the plant owner’s team. More specific examples
of this practice include:
• the purchase of integrated packages of equipment rather than
individual items;
• omission of drawings, such as piping isometrics or as-builts, which
become the responsibility of the fabricator;
• using lump-sum rather than per-item contracts for construction;
• imposing field measurement checking responsibility on the contractor;
and
• reducing office checks.
The list above is potentially endless. The same overall engineering
work is done, but the work is shifted to the direct field cost report.
Extra engineering work can be created by imposition of exacting design
and documentation standards, and reduced by the acceptance of previous
86 Handbook for Process Plant Project Engineers
Third Cycle
Conceptual Development
This page intentionally left blank
Studies and Proposals 91
Chapter 9
two parts: a pre-feasibility study with an accuracy1 of say ±20 per cent,
and a feasibility study with an accuracy of say ±10 per cent. This gives
the investor an opportunity to abort the study work or change direction
before too much money is spent.
The development of cost-effective designs is discussed in Chapter 11,
Value Engineering and Plant Optimization. Cost estimating is clearly at
the heart of study work; it is one of the investor’s main concerns. In
keeping with the need to carry out initial study work comparatively
quickly and less accurately, there are a variety of estimating techniques
available to estimate plant costs without doing too much design work.
Ultimately, however, an accurate estimate must be based on the submis-
sion of competitive bids for all items, and the bids must be based on the
adequate specification of equipment items, and adequate quantification
of sufficiently developed designs of other items.
9.2 Proposals
Studies may be regarded as a type of proposal, in that they relate to the
proposed construction of a process plant. However, we will restrict the
use of the term ‘proposal’ to that of an offer, namely an offer to build a
specified process plant at a certain price or price basis, a commitment
rather than an estimate. ‘Study’ will be used when an estimate of plant
costs, rather than a commitment, is submitted. The essential content of
a process plant project engineer’s work is the same for both, but there is
a fundamental difference in how the end product is used, in that the
relationship with the client is different. There is a corresponding
difference in the assessment of risk.
1
‘Accuracy’ is a word which is customarily used in this context, but the understanding
of what is meant varies quite widely; in fact the word is often used without any
understanding of its implications, such that the value of the quoted accuracy is
meaningless. The suggested usage is discussed in Section 9.3.
Studies and Proposals 95
these is the ‘curve price’, in which the entire plant cost is interpolated or
extrapolated from data on previous projects to construct similar plant,
the ‘curve’ in question being a graph of plant cost against capacity. Due
to limitations on similarity, it is seldom possible to assign an accuracy of
better than ±30 per cent to such an estimate.
The more developed techniques for preliminary estimates are based
on factorization. The basis for factorization is the observation that the
costs of component parts of process plants bear similar ratios to each
other in different plants.
Consider for instance the breakdown of the plant capital cost into the
following (all components include the associated site construction and
painting costs):
• civil works
• structural steelwork
• mechanical equipment
• electrical equipment and reticulation
• instrumentation and control gear
• piping
• transport to site
• indirect costs (engineering and management, insurance, etc.).
This breakdown includes all elements of the plant, that is to say that
possible other elements of breakdown (such as platework and valves)
are included in the above headings (for example mechanical equipment
and piping respectively).
Now the largest of these components is invariably the mechanical
equipment, and it is also the most fundamental component, being
arrived at directly from the process flowsheets and process requirements.
The other plant items follow from the mechanical equipment needs. So
in the simplest factorization technique, the plant cost is factorized from
the mechanical equipment cost, which is generally in the range of 30–45
per cent of the direct field cost, or 25–40 per cent of total cost, including
indirects.
There are many possibilities for improving the accuracy by designing
and estimating more of the other plant components, and factorizing
only the most intractable, say piping. The technique is also refined
by developing ratios of plant costs corresponding to each type
of mechanical equipment, for example the ratio of the cost of a
centrifugal pump to the cost of the associated civils, structurals, piping,
electrical, and instrumentation. An appropriate factor is developed
for each mechanical equipment type; the plant cost is the sum of
96 Handbook for Process Plant Project Engineers
9.4 Risk
The engineering and estimating work of a study or proposal is not
complete until its accuracy has been established. For pre-feasibility-type
work, where the level of commitment is relatively low ⫺ at most leading
to a decision to finance a full feasibility study ⫺ it is usually considered
acceptable to presume an accuracy based on experience; that is, based
on previous validated experience that enough design and costing work
has been done to justify the accuracy quoted, which is unlikely to be
better than ±20 per cent. It is also expected that any significant
uncertainties and hazards will be summarised in the pre-feasibility study
report, for detailed attention during the feasibility study.
However, when a firm commitment to build a plant is under consid-
eration, detailed assessment is normally required of all cost elements
and of everything that can go wrong and affect the viability of the
proposed commitment. The process is described as risk analysis. Many
treatises and methodologies are available, as befits the gravity of a
subject which can seriously affect the fortunes of major enterprises. We
will address the fundamental aspects.
A risk is defined as a possibility that a project outcome may differ
from the planned outcome.3 Risk can be quantified as a probability. The
2
Institution of Chemical Engineers and Association of Cost Engineers (UK): Guide to
Capital Cost Estimating, 2001.
3
There are other definitions, for example ‘the likelihood that an accident or damage will
occur’, which is more appropriate in a safety context. In the commercial context, there
is also risk of a better outcome (lower price, better performance), which must be
considered in quantitative decision-making.
98 Handbook for Process Plant Project Engineers
9.5 Accuracy
The ‘accuracy’ of an estimate is a widely misused and misunderstood
term. Mathematically, accuracy can be expressed by the limits of
a range of values within which the correct figure lies. If the only possible
solutions to a problem lie within the range 6–8, then the answer
may be expressed as 7 ± 1. However, life is not so simple for the plant
cost-estimator. His estimate is made up of a large number of cost
elements, none of which is certain, even as to the absolute limit of the
range of possible cost. There is no certainty in his life, only probability
and confidence. Certainty only exists when the project is over!
In fact, the estimate can best be presented as a probability curve, a
plot of estimated cost (x-axis) against the probability that the actual
cost will fall within a defined band (say ±2 per cent) of the estimated
cost (y-axis), which is shown in Fig. 9.1. The curve will clearly feature a
maximum value at the most probable cost, or the cost considered to be
most probable, and slope downwards on each side of the maximum.
That is about all that is known about the shape of the curve. Many
models (normal, skewed, Poisson, geometric distribution, etc.) are
assumed and used in practice.
Because it is a probability curve, sloping away on each side to
diminishing but never-zero values, it is only possible to assign limits
100 Handbook for Process Plant Project Engineers
quoted limits, he can at least be justified in saying: ‘Oh well, that was
the 5 per cent possibility’!5
9.6 Contingency
Here is another loosely used term. A recommended definition of
contingency, which comes from the American Association of Cost
Engineers, is:
A cost element of an estimate to cover a statistical probability of the
occurrence of unforeseeable elements of cost within the defined project
scope due to a combination of uncertainties, intangibles, and unfore-
seen/highly unlikely occurrences of future events, based on management
decision to assume certain risks (for the occurrence of those events).
This definition is rather a mouthful, and needs to be thought out
quite carefully. It may be easier to visualize in terms of the probability
curve mentioned above. A contingency is an amount to be added to the
estimated cost (assumed here to be the most probable cost, corresponding
to the top of the curve, but not necessarily so) to increase the confidence
level to an acceptable probability (say 90 per cent) of a cost that will not
be exceeded. In this definition, it is implied that a project has a fixed
scope, and any elements of approved scope change will be handled as
approved variations to the project budget.
This definition is not the universal usage of the word. In some quarters,
the ‘contingency’ is an amount included in the authorized project
budget to allow for any variations in scope or any lack of forethought
(or whatever) to ensure that the project budget is not exceeded; practi-
cally, a margin for error and for future changes ⫺ extra money in the
bank. This concept has in the past been taken to an extreme in certain
plush institutions, where quite large contingencies, even up to 20 per
cent, were routinely included in the budget of an authorized project by
the executive responsible. This contingency was not passed on to the
next level of authority ⫺ say the poor old project manager ⫺ thus it was
possible for the lower level to produce a mediocre performance in terms
of budget, while the upper level was a star. This is nice work if you can
get it ⫺ at the upper level!
5
OK, that is the second thing he will say, after the inevitable condemnation of the
project manager!
102 Handbook for Process Plant Project Engineers
6
Contingency calculation and management are often an exercise of power, rather than
mathematics. Some executives use the above techniques to eliminate all real contingency,
and then indulge in further power-play by requiring special authorization to access even
this bogus contingency.
Studies and Proposals 103
Chapter 10
10.1 Layout
Plant layout is the fundamental conceptual design activity which follows
on from the definition of the process and the mechanical equipment. It
was noted in the previous chapter, in Section 9.3, Estimating project
costs, that a preliminary plant cost could be factorized from mechanical
equipment costs with an accuracy expectation of ±25 per cent. Any
increased accuracy of costing depends primarily on determination of the
plant layout. Of course this may not have to be designed specifically;
but the applicability of similar plant layout designs, at least, needs to be
decided. The relevance of quoting the order of magnitude of costing
confidence is that it gives some idea of the influence of layout design on
plant cost.
Following on from the flowsheet, which determines the mechanical
equipment requirements, the layout reflects and determines most other
plant cost components, and its optimization is obviously critical to the
development of a cost-effective plant. It is equally critical to the
development of a functional and maintainable plant that is safe and
ergonomically acceptable, and can profoundly affect operating costs.
The terminology we are using is that, as was described in the basic
sequence planning of engineering work, ‘layout’ drawings are those
which are produced prior to the stage of receiving final equipment
details and structural design. The layouts therefore include overall or
critical dimensions only, whereas the final ‘general arrangements’ are
those used for controlling and verifying plant construction details.
Layouts are intermediate, not construction, drawings. This terminology
is not uniformly adopted in the process industry.
Layouts are developed by review and progressive refinement, often
considering many alternatives as befits their importance. There is no
106 Handbook for Process Plant Project Engineers
• The means of spillage removal (mainly for plants handling bulk solids).
• Facilities and access for major maintenance, including equipment
removal and replacement. These may include permanently installed
lifting devices (cranes, hoists, lifting beams, trolleys, etc.) and the
suitable configuration of surrounding plant and structures, for
example the provision of removable floor and roof sections, and
additional flanged connections in surrounding ducts.1
• Clearances and special requirements for operational hazards,
fire-fighting, and safety requirements.
• Unusual static and dynamic loads, and consequent structural
support requirements or isolation of vibrations.
• Noise attenuation.
• Any other requirements or advice from proprietary equipment
manufacturers, where applicable. (They should be a prime source of
information, comment, and eventual approval, although unfortunately
they cannot always be relied on to have good representation.)
It is necessary to check through the above list, at the very least, to
ensure that a proposed layout is appropriate for each piece of equipment.
In processes where materials are transported mainly in the fluid
phase, such as most oil refinery units, the piping design considerations
tend to dictate the layout, whereas for plants processing solids, such as
metallurgical reduction plants, the bulk solids handling design is the
major consideration.
The engineering of fluid handling systems, including piping and
ducting, and their influence on layout, is addressed in Chapter 14.
Piping considerations at the layout stage include not only the routing
(without creating plant access problems) but also support of the pipes,
access to valves, and provision for major expansion loops. Ducting,
particularly large diameter duct systems tied into a common stack, can
have a crucial influence on layout.
Bulk solids handling is addressed in Chapter 15. Some consideration
also needs to be given to the intermediate case of two-phase flow and
especially slurry transportation, and this is described in Chapter 16.
Bulk solids handling is inherently more complex and less flexible than
fluids handling, and has a correspondingly greater impact on layout
options. Conceptual engineering of the conveyors and gravity flow
1
It is easy to go too far with this, for example spending money on bolted connections,
which will hardly be used in practice (the ‘Meccano mentality’), when cutting and welding
may be acceptable.
108 Handbook for Process Plant Project Engineers
2
This subject is discussed further in Chapter 21, The Organization of Work.
112 Handbook for Process Plant Project Engineers
and a communication, which may be the model itself, the model with
notes added, drawings, or simply a set of instructions. It should be
noted that the communication aspect is an important criterion of
acceptability in its own right, and has to take into account the character
and needs of the people building the plant.
In Chapter 4, in the discussion on engineering co-ordination, we
highlighted the critical importance of the plant model as the instrument
by which co-ordination in space is maintained. A model which is clearly
and quickly understood by all disciplines (and by other reviewers, such
as the client) greatly improves design quality. There is nothing to beat
the ease of visualization of a physical model, but unfortunately it is rela-
tively expensive and slows down the design process. Advances made in
computer modelling in the last 30 years have taken us to the point where
the hardware and software has become almost as economical and effective
as the visionaries expected,3 and most design offices now utilize such
systems. If used by adequately skilled staff, they facilitate a process of
model creation which matches and becomes the layout, general arrange-
ment, and detail design process (see Fig. 10.3). Mutually consistent
drawings are automatically generated. There are also, of course, many
ranges of design and analysis software which integrate directly to the
modelling software. The trend of integrating associated design and
management software has continued to the point where integrated
software suites cover almost the entire engineering and management
development of a project.4
The further development and application of integrated computer-
assisted design and drafting is the way ahead, but the choices of
methodology to be made for a real, present-day project have to be
carefully considered, in particular the need and availability of suitably
qualified staff to handle the project’s peak load requirement.
Even with the clearest model presentation, there is a need for the
ongoing model development to be supervised by dedicated design
co-ordinators. The modelling software usually includes clash detection
and notification, but it is still necessary to ensure that in conflicts
the most important needs prevail and to maintain the integrity of
operational and maintenance access and ergonomics. There is also
frequently a need for ongoing independent review of the drawings or
output communication, from the point of view of their users.
3
Unfortunately, most of the visionary users lost their shirts in the process.
4
Creating more opportunities for visionary users to lose their shirts in the process! The
subject is discussed further in Chapter 28.
114 Handbook for Process Plant Project Engineers
Fig. 10.3 Virtual three-dimensioual model of plant
Plant Layout and Modelling 115
Fig. 10.4 Autogenous milling plant layout – showing
influence of bulk solids handling needs
This page intentionally left blank
Value Engineering and Plant Optimization 117
Chapter 11
1
If value engineering is required within a fixed-price framework, draw up a procedure of
how it will be done and specify what items and documents will be value-engineered, with
a specimen output report.
Value Engineering and Plant Optimization 119
Chapter 12
The actions are likely to include both design changes and operational
practices, which must be included in plant operation and maintenance
manuals, and operator training. A few examples follow.
• ‘None’ may bring up the possibility that a pump suction may run
dry, damaging the pump, and the recommended action may be to
install a low-level trip or alarm (or both) on the vessel from which
suction is drawn.
• ‘More of’ relates to quantities or properties of process substances,
and leads to consideration of excessive flow, over-pressure,
excessive temperature, oversized rocks, etc. with usually fairly clear
consequences and preventive action possibilities.
• ‘Less of’ could lead to recognizing the possibility of inadequate cooling
water flowrate to part of the plant when the flow is increased else-
where, with consequent damage to machinery; the solution could be
to install a flow-limitation device on the alternative user, or an alarm.
• ‘Part of’ is meant to include the possibility that the composition of a
material stream may vary, for instance that solids may settle out in
a pipeline that has a ‘dead leg’ under certain operating conditions,
with the consequence of blockage and unavailability of that part of
the line when it is needed. The recommended actions may include
the elimination of the dead leg by employing a recirculation system,
redesigning the dead leg to be entirely vertical so that settlement may
not occur, or simply prohibiting the operational mode in which the
problem can occur.
• ‘More than’ means more components are present in the system
than there should be, for instance the possibility that water may
be present in oil introduced into a hot vessel, with the consequence
of explosion. The recommended action may be to eliminate the
possibility of contamination by water at source, or to detect its
presence and initiate a shutdown system, or both.
• ‘Reverse’ means the opposite of the intended operation, for instance
reverse flow.
• ‘Other than’ looks for any abnormal operating conditions, other
than those already prompted.
We said above that it is necessary to look at every single pipeline,
item of plant, and device, and herein lies one of the sources of
potentially important omission. The definition of ‘device’ must
include anything that can affect the operation of the plant, including
power supplies and computers, and evaluating their participation and
response in operating deviations.
Hazards, Loss, and Safety 127
T
fdt = F ¥
2
Given the above, the probability of the unwanted incident occurring
will be D fdt events per year. The probability can be reduced by
installing additional safety devices, say a second safety valve or a pressure
switch which deactivates the source of pressure, but each of these will
have a fractional dead time, and the incident rate will be reduced to
D ¥ ( fdt )1 ¥ ( fdt )2 ¥
But it will never be zero. And with each extra safety device comes the
possibility of spurious action, upsetting plant operation when nothing is
wrong.
The second stage of Hazan is to evaluate the probable consequences
of the incident. Usually, these are related to the severity of the incident
(in the example of the pressure vessel, the maximum pressure experienced
in the over-pressure incident). The value chosen may influence both the
demand rate and the fractional dead time, so there may be some iteration
here. If we continue with the example of the pressure vessel, it is most
unlikely that it will explode if subjected to 25 per cent over-pressure
most codes require a hydrostatic test at around 50 per cent over-pressure.
There are all sorts of safety margins introduced into the various details
of the design but, then again, these are done for reasons which include
manufacturing uncertainty and service uncertainty (say, the inclusion of
a corrosion allowance). It becomes clear that the pressure at which a
vessel will fail is itself a matter of statistical probability. In practice, it is
frequently the case that an over-pressurized vessel or pipe fails at a
flanged joint, and the consequences (to people) are not so severe unless
the contents are flammable or toxic. For the given example, let us suppose
that we can expect the vessel to fail at a pressure of 170 per cent of the
rated pressure, that we have used this pressure in evaluating the demand
rate, and that a similar failure in the past has resulted in two fatalities.
Having done our best to quantify that there is, say, a 0.1 per cent
chance that a failure will occur in 100 years, and as a result an average
of two fatalities may arise in the ensuing explosion and fire, we come to
the final stage of evaluation, which is to decide whether that is accept-
able. This is commonly done by comparison with statistics that relate
to normal hazardous activities that most people are prepared to engage
in, like road or air travel. The designer may be able to demonstrate by
this means that the various features of hazard around a plant create an
130 Handbook for Process Plant Project Engineers
12.3 Commissioning
Refer to the notes included at the end of Chapter 24, Commissioning.
132 Handbook for Process Plant Project Engineers
Fourth Cycle
Engineering Development
and Detail
This page intentionally left blank
Specification, Selection, and Purchase 135
Chapter 13
13.1 Procurement
As we have seen from the sequential planning of engineering work,
not only are the project designs and their practical implementation
connected through the process of procurement, but also the flow of
information (both technical and commercial) back from vendors is
essential to keep the design process on track. There can be no possibility
of a successful project engineering effort without an equally successful
procurement effort; the two functions must be integrated from the basic
planning stage through to project completion. In Chapter 6 we
discussed some of the broad issues of procurement, in the following the
subject will be addressed at an operational level.
The engineering task that initiates the procurement process is the
defining of what has to be procured, by means of specifications, data
sheets, drawings, and work-execution plans, and this definition has
to be geared towards the capabilities of the suppliers, based on a
considered strategy and a knowledge of the marketplace. There remains
the work of:
• drawing up conditions of purchase and contract;
• packaging the commercial and technical content, and ensuring that
all aspects of the proposed agreements are covered;
• soliciting bids;
• receiving the bids systematically, and assessing them technically and
commercially;
• negotiating and finalizing agreements with selected bidders; and
• following up to ensure that the correct goods and services are
provided according to schedule, and that payment is made as
agreed.
136 Handbook for Process Plant Project Engineers
13.2 Specification
The word ‘specification’ has different meanings in different contexts,
especially in legal and patent usage. In engineering projects it means a
detailed description of the design and/or construction and/or perform-
ance of an item. The item can be an entire process plant, or one or a
group of pieces of equipment, activities, designs, or bulk commodities.
The usage of specifications is as old as the performance of projects, and
is the epitome of the statement made in the preceding review on project
management, Chapter 3, ‘Plan the work then work the plan’. First, you
decide exactly what you want (the specification), then you do it or get it.
The project design criteria document is thus the most basic of the
project specifications and, below that, any individual discipline detailed
design criteria and working practices, developed to guide the work of
the project in a manner acceptable to the overall objectives and/or the
client. There is some overlap in the meanings of ‘specification’ and
‘procedure’. Generally, ‘specification’ defines the product of an activity,
whereas ‘procedure’ defines how to get there. A procedure may be part
of a specification, and vice versa.
Here we are discussing procurement, and the specifications therefore
address the interface between the project organization and the supplier/
sub-contractor, in order to describe exactly what is required. In the
course of procurement, the specification is used in two contexts: firstly,
as the document which makes a competitive bidding process possible
and, secondly, as a reference for the work performance, describing
the products or services that are to be provided, and the standards of
acceptability.
is, the product that fulfils the required functions, with the required
reliability and durability, and at the minimum cost. The latter may be
adjusted to reflect the evaluated worth of performance-related features,
such as power consumption or maintenance costs, as discussed in
Chapter 8.
In order to obtain the best value, it seems sensible to produce a speci-
fication which reflects the minimum requirements that are acceptable,
thus enabling maximum competition and possibly innovative solutions.
In effect, this is a value-engineering approach. The specification is
aimed at identification of the essential requirements for performance
of the item, ‘performance’ meaning the definition of function and
how well, how economically, and how reliably the function must be
performed. Any non-essential requirements are identified and deleted.
This is one of the basic tenets of writing specifications, but limitations
to the practice include the following.
• If basic design choices are not made when detailed engineering is in
progress, it becomes impossible to complete the work. For instance,
with regard to equipment selection, once the layout has been frozen
(based on certain equipment types) it usually causes a lot of
disruption to change to other equipment types. The effect on the
layout, the occupation of space, the supporting steelwork design,
and on other disciplines just becomes too great and the potential
benefits are far outweighed by the cost of change.
• As regards bulk items (such as pipe fittings), standardization and all
the associated benefits will be lost if only functional requirements
are considered each time an order is made.
• Referring back to the discussion in Section 8.2 on lifecycle
considerations, we remarked on the difficulty of confirming maint-
enance cost and reliability claims made by equipment suppliers. We
concluded that it was often preferable to specify features which were
known to offer enhanced performance in these respects rather than
rely on competitive bidding to produce an optimal design.
The preparation of specifications is therefore a compromise between
getting the best value by restricting the requirements to the essential
performance of the item, and the sometimes conflicting needs of design
convergence, standardization, and inclusion of proven essential features.
All of these factors are important.
To some degree the compromise can be reached by using two stages
of specification. In the initial or study stage the specifications may be
almost exclusively performance-based, and serve to establish the basic
140 Handbook for Process Plant Project Engineers
the most part irrelevant, thus defeating the objectives of the previous
paragraph.
3
Minimum implies that the vendor is still responsible for the actual features employed,
provided that they are equal to or better than the minimum.
Specification, Selection, and Purchase 143
4
This is not universally applicable; not for instance when there are other overriding
laws, or when the purchaser has ‘qualified-out’ his obligation to inspect, or where there
is a warranty from the seller.
Specification, Selection, and Purchase 147
5
We have not discussed the use of second-hand equipment, which is sometimes the key
to project viability. One of the main consequences of buying used equipment is that the
vendor cannot easily give a serious guarantee, except by completely stripping and refur-
bishing the equipment at substantial cost (which is not known in advance). Thus the
onus of establishing fitness for purpose falls partially or wholly on the buyer, who must
tread warily. The situation is better if the used equipment belongs to the plant owner,
who is able to confirm its operational and maintenance history.
Specification, Selection, and Purchase 149
and needs relating to groups of items. There are many ways in which
this is practised; here is one example.
Group 1. These are critical items where the consequences of failure are
severe. An exercise of due diligence is necessary; that is, the intensity
of inspection should at least equal customary practice for comparable
work. Typically, all work should be carried out in accordance with an
appropriate code or standard, with an independent inspector to certify
compliance in accordance with an individual quality plan per item.
Group 2. These are items where compliance is required by statute,
for example pressure vessels. The applicable regulations dictate the
minimum requirements. If these items are also group 1 items, above-
statutory-minimum requirements may be needed.
Group 3. These are items where conditions of contract or commercial
policy dictate the levels of inspection, for example, in case the inspection
is linked to a final payment. There are no short-cuts here.
Group 4. These are the remaining items, where inspection is carried out
on a basis of economic evaluation,6 considering the selected supplier’s
reputation and facilities, the availability and credibility of conformance
certificates, the consequences of accepting non-compliant items, and
the cost of inspection.
For management purposes, the decisions on inspection intensity may be
classified into a few standard ‘levels’, for example no shop inspection,
random inspection of end product, and individual item quality plan.
Notwithstanding the above, most project inspection budgets are
under severe pressure, as rather a cynical view is taken on the value of
inspection, generally the result of unthinking criticism when defective
products pass inspection, without any understanding of the economic
impossibility of 100 per cent inspection. Possibly the most effective
countermeasure to the tendency to slash inspection budgets is to work
into conditions of purchase the requirement that suppliers and contractors
are obliged to pay for appropriate levels of additional inspection (by
third parties) in the case of failure of sample inspections. This may also
be the fairest and most effective option. Too often, in modern practice,
the quality supplier loses on price to those whose quality control is
purely cosmetic, and who ultimately (when quality failures are noticed,
too late) inflate the purchaser’s inspection costs while destroying his
inspector’s reputation.
6
In other words, the optimum input should be determined on the basis of
diminishing-return theory, as in Chapter 8 and Fig 8.1.
150 Handbook for Process Plant Project Engineers
13.7 Expediting
Project controls – the feedback loops of measurement and remedial
action to control performance according to plan – are customarily split
into cost, time, and quality functions. The time control system related to
procurement is customarily referred to as ‘expediting’, although the
term could be and sometimes is applied to any project activity (or the
perceived lack thereof), including engineering.
Inevitably, any activities directed towards the expediting of vendor
activities have an interface with other project activities which are related
by physical sequence or information flow – expediting is a two-way
activity. For instance, as regards the procurement of equipment, it is
also necessary to expedite:
• delivery to the vendor of any outstanding design information;
• approval, by the project team, of vendor drawings before manufacture
commences;
• concession requests (for deviation from the specification or from
approved drawings) as manufacturing difficulties arise;
• queries on the interpretation of drawings; and
• third-party inspection and witnessing of tests.
It is just as important to expedite the flow of information from the
project team to the vendor as it is to expedite the vendor’s own activities.
Many quality authorities argue that expediting must be separated
from inspection, in both line management and execution. No
one person should share both responsibilities and be exposed to the
ensuing conflict of objectives. Besides, the respective roles tend to
require different skills and personalities – the quintessential inspector
being a skilled and meticulous technician, while the expeditor may
be perceived as a bully who will accept no delays on any account.
Needless to say, this is an over-simplification which grossly undervalues
the maturity of some of the professionals involved. While the ideals
of role separation may be preserved with regard to more critical
items, the dictates of economy make it inevitable that there is a level
of criticality below which the roles of expediting and inspection are
combined.
For the most part, observations made in the preceding section about
the planning of inspection, may be considered interchangeable with
those on expediting, provided that quality is substituted by time. Both
functions require a plan as a basis of control, and a plan of execution
based on the criteria of criticality and limiting returns. And both
Specification, Selection, and Purchase 151
Chapter 14
Fluid Transport
p
h=
rg
1
Any consistent units may be used; for example, in this case h is in metres, p is in
Pascals (N/m2), r is in kg/m3, and g is in m/s2.
2
A classic reference is A. J. Stepanoff, Centrifugal and Axial Flow Pumps: Theory, Design
and Application, 1992.
Fluid Transport 155
h = k1 2 = k1π2 n2 D2
φ
k1 =
g
Qp Qhrg hmg
P= = =
E E E
where E is the efficiency of the pump, and m is the mass flowrate. The
efficiency versus flowrate curve generally has a maximum at the pump
design flowrate, and curves downwards at lesser and higher flowrates.
The power-versus-flowrate curve can be deduced from the head-
versus-flowrate and efficiency-versus-flowrate curves for a given fluid
density, using the equation above. It is found that the power curve tends
to rise continuously as flowrate increases for a radially vaned impeller,
while for a backward-sloped impeller the power rises less steeply (due to
the head decrease) and may reach a maximum value and then decrease,
which is described as a ‘non-overloading’ characteristic. The curve
shapes are summarized in Fig. 14.1.
We will turn now to the piping system. Pipelines should theoretically
be sized by a process of economic optimization, balancing the increased
capital costs of greater pipe diameter against the increased pumping
costs of reduced diameter and consequently higher head. Considerations
of heat loss, abrasion, and pump cavitation may sometimes also affect
the optimization process. In practice, it is usual to select as design basis
a range of standard pipeline velocities, for example 1.0–2.0 m/s for
Fluid Transport 157
pump discharge lines of increasing diameter, and 0.5–1.0 m/s for suction
lines. The economics of such design criteria should be reviewed
for the circumstances of individual major projects, especially where
construction materials or power costs are unusually high or low, and
where individual lines are very long or expensive. The design basis
should be qualified in terms of whether the flowrate used is the normal
or maximum value, and the implications taken into account when
setting up the standards for the project. We will assume in the following
that the design is based on maximum flowrate.
The dynamic (or frictional) head loss hf may be calculated from
V2 l
hf = 2 f ¥ ¥
g d
where f is the Fanning friction factor for the Reynolds number and the
pipewall roughness, from standard hydraulic charts (for example from
Perry’s Chemical Engineer’s Handbook). (Note: we are using the
Fanning factor because we have quoted Perry as a reference manual. Be
careful: Darcy’s factor is also commonly used – this is four times the
Fanning factor, so the formula must be adjusted accordingly). V is the
mean fluid velocity in the pipe of length l and diameter d. For a given
flowrate, V is inversely proportional to the square of the pipe diameter.
Allowing for the decrease in friction factor as the pipe diameter
increases, approximately
k2
hf =
d5
Given this exponential relationship between head loss and diameter, the
advantages of rounding up to the next pipe size when in doubt should
be obvious, as well as the severe penalty of undersizing the pipe.
The majority of process plant pipelines operate at moderate pressures,
say less than 10 bar. Consequently, as can be verified by simple
hoop-stress calculations, for steel pipes the pressure-induced stresses are
fairly nominal in relation to the material strength, and it is not necessary
to increase the thickness much as pipe diameters increase. Pipe wall
thickness is selected to provide adequate corrosion allowance and
mechanical strength and rigidity, particularly in bending. This is not the
case for plastic pipes, where more care must be taken to avoid pressure-
induced failure, and it becomes more important to understand system
characteristics by which over-pressure may arise.
158 Handbook for Process Plant Project Engineers
h = hs + hd
hd = k3Q2
where k3 is established from the dynamic head calculated for the
flowrate for the defined normal duty.
The system head envelope for different static heads (and, if required,
different control valve settings) can now be plotted against the flowrate,
Visualizing heads
The calculation of fluid flow and pump performance is usually carried
out in modern design offices with a computer, and many programs of
varied sophistication are available. However, reliance on computers
may be accompanied by a lack of understanding of what is really going
on, and with this lack of understanding, intuitive knowledge on system
design may also be lost. It also becomes more possible to make an
order-of-magnitude error, if computer input is incorrect!
Fluid Transport 161
hf = 2 f ¥ (V 2 Ⲑg ) ¥ ( L Ⲑd )
= 2 ¥ 0.006 ¥ (22 Ⲑ 9.81) ¥ ( L Ⲑ 0.25) ⬇ 0.02 ¥ L
If the static head is 20 metres, and the pipeline is 100 metres long, one
would expect a total system head requirement in the order of 25 metres,
so this should be well within the range of available centrifugal pumps,
without resorting to 2-pole speed.
14.2 Gases
For the most part, the design of gas transport systems follows the same
lines as for liquids, except for the phenomenon of compressibility and
consequent density variation. System performance is presented on a
head basis in the case of turbo-compressors or fans, which corresponds
to centrifugal and axial flow pumps, in the same way as for pumps.
There are applications, such as most flue gas or ventilation systems,
where the density variation is slight, and the same equations and design
systems can be used as for liquids, except that economical pipeline
design velocities are usually in the range of 10–30 m/s, the higher values
being for higher diameters and lower gas densities.
When the effects of density variation cannot be ignored, for example
with pressure ratios (maximum to minimum within the system) of
greater than 1.2:1, calculation of the head requires the assumption of a
relationship between the pressure and density (or its inverse, specific
volume ) of the gas. The two most widely used expressions are the
‘polytropic’ relationship
p n
= constant
n -1
È ˘
Ê n ˆ Í Ê p2 ˆ n ˙
h=Á RT - 1
Ë n - 1˜¯ 1 ÍÁ p ˜
Ë 1¯ ˙
ÍÎ ˙˚
h( J / kg )
h( metres ) = ⬇ 100h(kJ / kg )
9.81( metres / s2 )
h( metres )mg h( J / kg )m
P= =
E E
where P is the power, m is the mass flowrate, g is the acceleration due to
gravity, and E is the efficiency.
A further useful relationship is
n - 1 (γ - 1)
=
n γE
from which it is seen that n and c are equal when the efficiency is unity,
as one would expect. The head equation is in fact not very sensitive to
the value of n. When calculating the head required for a typical process
application, it usually suffices to take E as 0.7, at least for a first itera-
tion, unless there is better knowledge of the efficiency of the intended
compressor.
Example
For an air compressor, the operating conditions are
Inlet pressure = 1 bar absolute
Discharge pressure = 8 bar absolute
Inlet temperature = 20 °C = 293 K
Flowrate = 300 Nm3/h ≈ 360 kg/h = 0.1 kg/s
164 Handbook for Process Plant Project Engineers
Evaluate the polytropic head and power absorption, assuming that the
polytropic efficiency is 70 per cent.
Solution: for air, R = 0.29 kJ/kg K and c = 1.4, so
Ê n - 1ˆ Ê (γ - 1) ˆ Ê 1.4 - 1 ˆ
ÁË n ˜¯ = ÁË γ E ˜¯ = ÁË 1.4 ¥ 0.7 ˜¯ = 0.4
0.4
0.29 ¥ 293 ÈÊ 8 ˆ ˘
h= ÍÁ ˜ - 1˙ = 276 kJ / kg (or 27 600 m)
0.4 ÍÎË 1 ¯ ˙˚
276 ¥ 0.1
Power, P = = 40 kW
0.7
gh 9.81 ¥ 3000
= / 0.5 = / 0.5
= 250 m / s
Fig. 14.3 Centrifugal compressor installation. Note that the designer has
chosen to locate the connecting piperack on a level above the
compressor; this provides the facility to anchor the piping loads
caused by thermal expansion, and also maximizes access to
the compressor (while minimizing the elevation). Note also the
bypass control valve for surge protection
is unstable and surging will be experienced, that is, periodic flow oscilla-
tions. These are liable to cause stress and overheating, which the
compressor may not be able to withstand, and which are anyway not
usually tolerable for process performance. It is essential to check that
the minimum flow requirement of such machines does not overlap the
range of process operation. It is possible to protect the machines by
installing bypass systems coupled to minimum flow detectors ⫺ these
often come as part of the package ⫺ but the protection systems may be
relatively expensive, and should be taken into consideration at the time
of purchase (not when the compressor has been reduced to a pile of
scrap!).
In general, suitable experience should be brought to bear when
ordering compressors and pumps, especially for expensive applications.
Considerations should include the following.
• Suitability for the full range of process duties. This includes staying
clear of compressor surge, as mentioned above, and also rotational
speeds which are too close to critical values.3
• Potential start-up and shutdown problems.
• Efficiency and power cost, at normal duty and other required duties.
• Adequacy of shaft sealing arrangement and other rubbing seals where
applicable.
• Facilities required and available for cooling.
• Capacity control.
• Adequacy of instrumentation and protective devices.
• General maintenance requirements and reliability in relation to process
duty; possible need for an installed standby unit.
• Availability and cost of spare parts and service.
• Standardization, where possible, of components such as shaft
couplings, bearings instrumentation, and seals. (Shaft seals are the
most frequent maintenance item on a pump.)
• Above all, the track record of the application of similar machines to
the process duty in question, and the track record of the machine
vendor under consideration.
API (American Petroleum Institute) standards are commonly
referenced for the purchase of compressors and pumps for oil refineries
3
In the author’s experience, failure by process technologists to specify the full range of
operation is the most frequent cause of the selection of inappropriate equipment,
especially for turbo-compressors. The duty specifications should be proactively cross-
examined.
Fluid Transport 167
and demanding process duties. However, these standards are often too
demanding for light-duty applications, such as pumping water or
compressing air, and non-continuous duty. There are many national
industrial standards applicable to ‘chemical’, water, and air or flue-gas
duties. Engineers unfamiliar with these may be better advised to adopt
simple functional specifications, coupled with the input of adequate and
relevant experience, to evaluate what is offered. API standards are not
at all applicable to most metallurgical applications, which centre on
slurry pumping, and for which vendor standards often suffice.
5
The polytropic relationship is T2 /T1 = (p2/p1)(n⫺1/n).
This page intentionally left blank
Bulk Solids Transport 173
Chapter 15
15.2 Conveyors
Devices which move bulk solids may be roughly classified into two
groups: those which transport material in discrete quantities by
independently moveable containers, and those which convey a continu-
ous stream of material. In general the first class is appropriate for lower
capacities and higher material lump sizes, whereas the second class is
preferred, wherever reasonably possible, for process plant applications
because of compatibility with a continuous process.
We will therefore not dwell on containerized transport devices.
The equipment types include wheelbarrows, skips, ladles, railtrucks,
roadtrucks, overhead cableways, etc., and their usage is more often at
the feed or product end of the process plant, or for collecting spillage.
The most widely preferred device for continuous material conveying
is the troughed belt conveyor, its advantages being reliability, ease of
maintenance, high energy efficiency, and, except for very short distances
(say, less than 15 m), relative economy. There is a correspondingly
voluminous quantity of literature available on its design and application.
The Conveyor Equipment Manufacturers’ Association of USA (CEMA)
guides are recommended for general reference, being generally available,
widely used, and well presented.
The CEMA guides can readily be used to establish the design parameters
of most in-plant conveyors, and conveyor belt suppliers are usually
happy to check and comment on the calculations, as a service. Longer
or higher-power-draw conveyors, for example over 1 km or 250 kW,
may require more specialist design attention because of dynamic effects
relating to the elasticity of the belt.
The basic methodology for conveyor design follows.
1. Select the belt width and load cross-section formed by the troughing
idlers. This choice is based on the characteristics of the material
conveyed, the required maximum capacity, and the belt speed.
2. Calculate the power absorbed, this being the sum of power to elevate
176 Handbook for Process Plant Project Engineers
There are several other types of conveyor, many of which have been
found to satisfy a particular niche, and are often incorporated in a
particular process technology package. There is one other type, the
screw conveyor, which requires special mention because of widespread
and traditional usage. Advantages of screw conveyors include:
• relative simplicity of design, manufacture, and maintenance;
• ease of dust-tight enclosure;
• facility to integrally include a feed extraction facility at the back end
of the conveying section;
• possibility of providing multiple feed and discharge points on one
conveyor; and
• The availability of ample information on many types of application,
and the corresponding recommended design variations.
Features whose design may be varied to suit the application include the
flight construction, percentage fill, speed, and design of intermediate
hanger bearings (whose need may also be eliminated by appropriate
shaft diameter and length).
There is a CEMA guide for screw conveyors, and much other useful
literature, especially from manufacturers. A cautionary note is appro-
priate: whereas screw conveyors have been found to be the preferred
choice for many applications, their choice has also been the downfall of
several plants. Typical problems include the following.
• Application on an unsuitable material. For example: too sticky,
or building up on the flights and shaft rather than flowing over
them; too abrasive; or too lumpy, jamming between flights and
trough.
• Failure to adequately de-rate the capacity and increase the power
for upward inclination. The basic flow mechanism within a screw
conveyor is that of material sliding down an advancing inclined
plane; therefore it can be appreciated that the capacity decreases
asymptotically as the inclination approaches the helix angle.
(It is possible to improve the performance at upward inclination,
including vertically up, by designing the conveyor as a fully flooded
Archimedean screw, but the design basis is different and the power
draw higher).
• Use of unsuitable intermediate (hanger) bearings in an abrasive
application.
• Underestimation of the power requirement of the feeder section; this
is addressed below.
Bulk Solids Transport 181
15.3 Feeders
Feeders interface to the gravity flow of hoppers and silos, and promote
and control the passage of materials onto conveyors and containers or
into process equipment. There are very many types which have found
application for different materials, material lump sizes, capacities,
control requirements, and hopper or silo configuration. The most
popular types include vibrating feeders, belt feeders, screw feeders,
apron feeders, and rotary or star feeders. It is usually of critical
importance to ensure that the type of feeder selected has a proven
history of suitability for the application.
We will not consider the merits and details of various feeder designs:
we will rather address the nature of the interface between the feeder and
the bin of material above it, which impacts on the selection and design
of the feeder, and is too frequently the cause of design error.
The feeder has to function with a volume of material in the bin above
it. Obviously, if there is not always a positive volume of material it cannot
produce a controlled output, and it is therefore not a feeder. Usually,
the bin has a converging flowpath, say a cone, where it joins onto the
feeder. The geometry, construction, and surface roughness of this
flowpath affects the rate at which material can flow into the feeder, the
vertical load transmitted onto the feeder, and the power required to
drive the feeder. In the converging flowpath, frictional forces oppose the
oncoming load of material, relieving the pressure applied onto the feeder.
The bin, or equally the flowpath down into the extraction point below
a stockpile, must be designed to provide the required amount of ‘live’
(that is, accessible) stored material, without blockage. It is obviously
unacceptable, and yet not unknown, for large silos or stockpile systems
involving substantial investment to be useless due to susceptibility to
blocking. Blockage is invariably associated with the design of the bin
bottom and outlet. Obviously the outlet has to be large enough in
relation to the material lump size, for example at least three times the
greatest linear dimension. Equally, the slope towards the outlet must be
sufficient to induce flow, although this may be promoted by the use of
low-friction liners, vibrators,1 fluidizing air (for fine materials), or
airblast devices. A high slope-angle, say 70° or greater, may ensure the
flow of all but the stickiest materials, but has expensive consequences to
the cost of the retaining structure per unit volume of stored material: it
1
Caution: in an unsuitable application (generally, with fines), vibrators may cause
consolidation of the material, leading to blockage.
182 Handbook for Process Plant Project Engineers
pushes up the height. An angle which is too low will result at least in
dead material at the bottom of the bin, in a funnel-flow pattern which
restricts the maximum withdrawal rate; it may also lead to the forma-
tion of a stable ‘rathole’ above the outlet, effectively blocking it.
Several practising bulk solids flow consultants have evolved systems
to calculate appropriate bin and flowpath geometry, and selection
of liner material, based on the measured properties of samples of the
material handled (in particular the angles of internal and external
friction). Such rationally designed systems also embrace the feeder
interface, and generate design parameters which permit the confident
selection of the right feeder. The accuracy of the calculations performed
is obviously no better than the degree to which the samples tested
are representative of eventual service conditions. The possibilities of
variable material and moisture conditions, material consolidation over
time, and service deterioration of liners must be considered.
In the case of major storage facilities, it is generally agreed that the
capital expenditure involved warrants the cost of substantial testwork
and consulting fees. For smaller installations, for example, hoppers
containing say 20 m3 of stored material, it is common practice for the
experienced engineer to design the bin outlet geometry on rules of
thumb related to previous experience with similar material. For
instance, the following may be considered appropriate as a design basis
when a similar reference application exists to verify it.
1. Minimum bin opening (width of slot) = the greater of 3 maximum
lump size (3 50 = 150 mm in this case) or the minimum successfully
proven previously, which is 250 mm in this case.
2. Minimum valley angle in convergent section = 60°. Note: the valley
angle is the true angle to the vertical of the line of intersection
between two adjacent sides.
3. Assume that the downward load onto the feeder is equivalent to the
weight of a volume of material equal to the area of the bin opening
the width of the feeder inlet slot (250 mm). This implies that the
effective height of material equals the slot width: although there
may be, say, 6 m of material above the opening, the load has been
greatly reduced by the converging flowpath. This is a conservative
value the actual load is usually lower.
4. Assume that the maximum force to shear the material at the feeder
interface equals the load due to material weight as calculated in
point 3. (In other words, the equivalent coefficient of internal friction
is 1, which is a conservative value that may be required at start-up;
the actual running value should be lower.) This is the effective belt
Bulk Solids Transport 183
Chapter 16
certain size, shape, and velocity) seem to bounce off it. However, larger
and sharper particles may quickly tear and destroy the rubber. Quartz
and metallurgical ores tend to be highly abrasive, while many chemical
crystals are relatively unabrasive and can be conveyed at velocities
where no liner is required.
Abrasion is usually a major consideration in pump selection. Except
for the mild chemical-type duties already referred to, slurry pumps are
of special construction, essentially to resist abrasion to the pump and its
shaft seals but often also to permit fast maintenance and liner or com-
ponent replacement for very abrasive duties. Generally either replaceable
rubber lining or specially hard abrasion-resistant alloys are used; shaft
seals are of special design, and are usually continually flushed with
clean water. Impeller tip velocities are kept within proven limits for the
application. Power transmission is often through V-belts rather than
direct-drive couplings, to permit exactly the required speed to be obtained
(rather than synchronous speeds) and to permit speed optimization in
service (varying the impeller diameter is not so convenient).
There is a large volume of literature available on slurry system design,
including formulae for critical velocity (the most popular is the Durand
equation; refer to Perry, for details see p. 157) and for more precise flow
resistance determination. These have their limitations however: design
is best when based on experience of the duty. Much information can be
obtained from the major slurry pump manufacturers, including
Envirotech (Ash Pumps) and Warman. For critical applications,
for example long tailings lines, it is wise to arrange for tests to be carried
out on samples of the actual slurry, ultimately in closed circuit
pumping loops, with interpretation and system design by an experienced
consultant.
We stated above that for most applications the slurry behaves as a
fluid of modified density, and that viscosity effects can usually be
neglected. However, there is a limit to the permissible solids concentra-
tion. Relatively coarse solids can simply not be carried in suspension
above a certain solids concentration for the particular slurry, say 40 per
cent by mass. In the case of very fine solids, a ‘thixotropic’ mixture is
reached at higher solids concentration, in which the solid particles stay
in suspension but the slurry viscosity is markedly increased. In fact
the viscosity no longer exhibits a Newtonian relationship of direct
proportionality between shear stress and velocity gradient. This is an
area to keep away from in normal plant design, but for special applica-
tions like mine backfill plants, when high solids concentration is critical,
it may be desirable to commission a special design based on materials
188 Handbook for Process Plant Project Engineers
Valves for slurry duties are of special design. Control valves and
check valves are not normally used, although some special valves (such
as pinch valves) are available. Isolation valves have to be designed to
avoid build-up of solids. Knife gate valves (in which the gate is designed
to cut through solids), rubber diaphragm valves, ball valves, and plug
valves are commonly employed; it is necessary for the engineer to verify
the valve’s application by relevant experience, just as when purchasing
process machinery.
If a slurry stream flowrate has to be controlled, an attractive method
is to use a variable speed pump. If the flow has to be divided the most
usual system is to employ an elevated tank for the purpose, and control
the divided flows with variable overflow weirs.
Returning to the subject of pipeline routing, we have observed that
some slurries are more forgiving than others when subjected to a
convoluted flowpath. It is essential to have a good understanding of
these characteristics when laying out the pipelines. If in any doubt,
assume that it is necessary to restrict the number of bends to a
minimum, and keep their spacing far apart. Bends are subject to extra
abrasion, apart from causing plugging of the line. In any event, long-
radius bends should be used, and 45° entries should be used where flows
converge.
It may not be easy to comply with the need for elimination of bends
without creating serious access obstructions within the plant. When
laying out a plant employing slurry transportation it is advisable to
include the routing of the slurry lines, and certainly the more important
lines, as part of the conceptual design. Verifying that these lines are as
short and straight as possible should be a prerequisite to acceptance of
the overall layout for detailed design. These lines may in practice be the
most significant aspect of plant reliability.
Flushing facilities, for use when closing lines down and for blockage
removal, are required for all slurry systems. Stagnant areas and unne-
cessary turbulence must be avoided. Pump suction lines require special
care, in order to be kept as short and straight as possible and without
excessive diameter.
• The adequacy of the vertical shaft diameter. This depends not only
on the torque, which is easily calculated, but also on the required
bending moment, which is not. The bending moment is caused by
eddying and unstable flow within the tank; thus it can be promoted
by insufficient size or number of baffles. Research articles have been
published on the subject, but it is more practical to be guided by
relevant experience.
• The impeller mounting and maintainability.
• The materials selection and corrosion and/or abrasion protection of
wetted parts.
We stated above that ‘usually’ an agitator is used for maintaining
slurries in suspension, but this is not always the case. Agitators can
sometimes be eliminated by careful tank design in relation to the
flowrate, and it is always desirable to eliminate equipment items when
possible; this is not just to save initial cost – if the item does not exist it
will not fail. Typically, pump suction tanks (‘sumps’) are sized to avoid
the use of agitators, making use of tank bottoms acutely sloping
towards the outlet. The size and design of these has to be carefully
judged against previous applications for the particular slurry and
comparable minimum flowrate. It is also often possible, with similar
care, to design collecting and transfer tanks to function without
agitators. If in doubt about the application, it may be instructive to
compute the power input P to the tank in terms of the fluid head
V2
P= Qr + hgQr
2
where V is the fluid velocity in the pipeline, h is the fluid impact height
(that is, the height of the pipe centreline above the surface in the tank),
Q is the slurry flowrate, and r its density. The resulting power input, per
unit volume, may be compared with the corresponding figure based on
absorbed shaft power per unit volume for similar successful agitated
tank applications. The energy of the incoming slurry is not as well
utilized as in the case of an agitator, so it is wise to look for about
double the latter figure to compensate. A worked example is included in
the next chapter.
π
Time taken for change = distance traversed/speed = (3d ) /V
4
Dividing and simplifying
1
Force = momentum change/time = dl rV 2
3
π
This will clearly not hold for l > length of the bend (3d ), so if we take
4
π
l= (3 ¥ 0.2) = 0.47 m
4
1 kg m2
Force = ¥ 0.2 m ¥ 0.47 m ¥ 500 3 ¥ 202 2 = 6 kN
3 m s
The author’s belief, based on observation of some failures, is that in
fact even greater forces can be generated. Factors which may magnify
the reactions include:
• pressure pulsations in the pipeline;
• velocity pulsations; and
• dynamic response of the pipe and support structure to impact
loading and periodic oscillation.
It may be advisable (especially for pneumatic conveying) to design for
a support reaction equal to the product of the cross-sectional area of the
pipe and the maximum pressure. For instance, in the case of the 200 mm
line above, if the initial pressure of the motive air is 2 bar gauge
(200 kN/m2), the force is
π
¥ 0.22 ¥ 200 = 6.3 kN
4
always a two-phase flow. The velocities are kept much lower than for
pneumatic conveying, but the forces can still be large enough to cause
significant oscillations in the furnace outlet bends. The piping here is
invariably designed to be rather flexible (and therefore comparatively
flimsy) to accommodate expansion. In this case robustly designed
dampers may be considered, to allow expansion but prevent oscillation
of the pipes. Pipe flanges, which are usually a necessity for equipment
isolation, should be located in a position to minimize the bending
moment that will be experienced if the pipe oscillates. Usually, the best
position is close to the bend.
Hydraulic Design and Plant Drainage 195
Chapter 17
flow area
R=
wetted perimeter
The Reynolds number Re, for the minimum velocity, and based on
kinematic viscosity v of 1.0 × 10⫺6 m2/s is
1 Ê h ˆ gde
f = 0.004 = Á ˜
2Ë l ¯ V2
h 2 f V 2 2 ¥ 0.004 ¥ 2.52
=s= = = 0.005
l gde 9.81 ¥ 1.0
contain such a high fines content as to modify its viscosity. In this case
the usual practice is to use the Manning formula
2 1
V = 1 R3 s2
n
or
2 2
s = n V4
R3
The Manning formula can be derived from the Chezy formula by
1
assuming that f is inversely proportional to R 3 . The roughness factor n
can be taken as 0.012 for reasonably smooth steel pipes when SI units
are used. For the case above
0.0122 ¥ 2.52
s= = 0.56 %
0.251.33
Note that in units of feet and seconds, the Manning formula is
expressed as
2 1
1.486R 3 s 2
V =
n
in which case n is the same as in SI units. However, some textbooks
incorporate the constant 1.486 into n, in which case n is different by this
ratio. (1.486 is the cube root of the number of feet in a meter.) Clearly,
one has to be careful when using tables for n. In any event, the values of
n and f (or the roughness ratio from which f is selected) are not very
accurately known for most practical cases.
We will now concentrate our attention on water-borne slurry applica-
tions; single-phase gravity flow applications are relatively simple to
design, and seldom cause problems that are not fairly obvious by
application of common sense. The challenge in the case of slurry
applications is to avoid blockage by settlement of the solids (calling for
high enough velocities), and to minimize abrasion and spillage (calling
for low velocities). All of these concerns are assisted by employing
relatively straight flowpaths with minimal obstruction.
Often, the design of a plant utilizing slurry-phase transport is based
on compromise between, on the one hand, building process stages at
198 Handbook for Process Plant Project Engineers
1
This would be for a duty such as flow of crystals within a mother-liquor in a chemical
plant. Applications such as metallurgical plant slurries tend to require higher velocities
because of higher specific gravity difference between the solids and the liquid.
Hydraulic Design and Plant Drainage 199
1 ¥ 0.25
R= = 0.167 m
1.0 + 2 ¥ 0.25
purpose we have noted that in a similar application the slurry was kept
in suspension in a box of 4 m3 volume, with a hydraulic input power of
2.2 kW. In this case the process parameters we will use are: velocity V at
pipe outlet = 3 m/s, minimum flowrate Q = 0.25 m3/s, and slurry density
r = 1500 kg/m3, and we have guessed that the pipe outlet may be a
distance of 1.0 m above the splitter box water level. Using the formula
from p. 191, we see that the hydraulic power input
V2 32
P= Qr + hgQr = ¥ 0.25 ¥ 1500 + 1.0 ¥ 9.81 ¥ 0.25 ¥ 1500
2 2
= 4500 W = 4.5 kW
As our Fig. 17.1 splitter box has a diameter D of 2.0 m and the height to
the water level h is the same, the contained volume is 6.3 m3 and the
power per cubic metre is 4.5/6.3 = 0.7 kW. This exceeds the value of
0.55 kW for the existing tank, so there should be no settlement, and the
margin of excess is not too great, so there should not be excessive
splashing. We include also the part of the plant hydraulic diagram for
this section in Fig. 17.2. (Note: these diagrams can be and are presented
Chapter 18
Observations on Multi-Discipline
Engineering
each application, and the basic pattern of column spacings and floor
elevations adopted (in conjunction with the layout designer). At this
stage there must be a critical review of design for horizontal stability.
The most economic steel frame design invariably incorporates many
braces, but these obstruct the placement of large ducts and pipes and
do not accommodate the passage of cranes and plant operators. It is
critical to develop basic rules in each building, at the conceptual design
stage, as to which bays may accommodate what type of bracing, in
accordance with the needs of:
• plant access and maintainability;
• constructability; and
• space needed for ducts, pipelines, and cableracks.
Space reserved for bracing should be shown clearly on conceptual
drawings. Portalized structures should be considered when there is
doubt about conflicting plant space requirements, or they may be
dictated by crane travel path.
When it is decided to enclose any space, the basic concept for the
ventilation system must be determined in so far as it affects the
structural design (usually, the type of roof). Note that the ventilation
requirements are very much process-related.
the points of issue within a process plant design remain the same; we
will try to summarize the most persistent of these, most of which relate
to inter-disciplinary information flow and co-ordination.
The plant instrumentation and control requirements are directly
linked to the process design as portrayed on the P&I diagrams. There
is therefore a natural direct route of communication between the
process engineers and the instrument engineers regarding the overall
measurement and control system functional design (the ‘control
philosophy’) and the individual instrument data sheets describing the
process conditions and measurement range and accuracy required.
However, the instrumentation and control systems also serve mechanical
equipment in non-process-related functions, such as machine protection
and control of machine auxiliary services. This latter requirement,
dependent for the most part on information from mechanical engineers
and machine suppliers, must also receive adequate attention when
scoping instrumentation and control work and devising co-ordination
procedures.
The instrument/mechanical interface is not just a matter of ensuring
that machine instrumentation is accommodated within the overall
control system. It is necessary to work in the opposite direction, to
ensure that all instrumentation and control features of machinery
are fully compatible with the overall plant systems, including their
functionality, interface parameters (such as signal and electric power
characteristics), and component standardization. This requires a direct
participation of the instrument engineer in the specification and
selection of most mechanical equipment packages. The instrument
engineer must develop a complete understanding of the functioning
of the mechanical plant and its critical needs, for example, signals
which are essential to plant safety and must be configured in a ‘fail-safe’
manner.
We previously mentioned the importance of co-ordination of cable
and cablerack design with electrical piping and structural design, with
the main objectives of preservation of plant access, avoidance of clashes,
and the integrated design of supports. ‘Clashes’ in the instrumentation
context should include any possible electromagnetic interference due
to the proximity of high-current cables and magnetic fields. It is also
important to identify the space, support, and access needs for local
instrument panels, marshalling boxes, and the like, and include them in
integrated layout development work just like an item of mechanical
equipment. Failure to give these items adequate attention frequently
results in a ‘retrofit’ design, whereby instrument panels obstruct access
214 Handbook for Process Plant Project Engineers
Chapter 19
19.1 Structures
As stated in the previous chapter, it is preferable to design all elevated
equipment supports, plant enclosures, access platforms, piperacks, and
pipe and cablerack supports as steel structures. Exceptions typically
include:
• supports for major machinery items, where heavy dynamic loads
make it advantageous to mount the items on elevated, heavy
concrete structures or blocks integral with the foundations;
• areas where the need for fireproofing makes steel structures less
attractive;
• areas where elevated concrete flooring is needed for collection and
disposal of liquid and solid spillage, or for protection from fires
beneath;
• areas where reinforced concrete is preferred for corrosion resistance
(sometimes unjustifiably!); and
• areas, such as electrolysis houses, where some degree of electrical
isolation is required.
For the rest, even where reinforced concrete construction may at first
appear to be a more economical option, it is usually avoided because of
the relative difficulty of modifications and the increased quantity of site
work: steelwork fabrication can proceed in the shop while foundations
are constructed.
Steelwork is designed in the format of line diagrams, based on the
layout drawings and equipment loadings. From these, steelwork
arrangement and detail drawings are prepared, showing also the detailed
connections to adjoining parts (column bases, supported equipment,
etc.); the arrangement of flooring, handrails, kickflats, stairs, building
cladding, and the associated purlins, and architectural details. At this
point the drawings are typically passed to a steel detailer who is
employed by the steelwork fabricator. The detailer designs individual
connections, for example welded gussets and bolted joints, and produces
detailed manufacturing drawings for each steel member, bulk material
cutting lists and diagrams, and a corresponding numbered assembly
drawing to facilitate manufacturing control and erection. The responsible
structural engineer checks that the details are structurally acceptable;
invariably they have to conform to pre-ordained typical details.
Sometimes the workshop detailing is carried out by the project
steelwork design team or a separate sub-contractor, either because such
a work breakdown is preferred or because the steelwork fabricator does
Detail Design and Drafting 219
adjustment. The end result may look as if the intermediate frame could
have been designed as part of the conveyor headframe, but the
mental liberation provided by regarding the intermediate frame as
a separate entity seems to have a positive benefit for both design and
construction.
19.2 Piping
In Chapter 14 we dealt with the fundamental aspects of piping
engineering, by which are set up a system of pipework specifications
which identify all the components (pipe, fittings, valves) within
each pipeline. We also discussed the need for basic design criteria
which included the means of dealing with thermal expansion, the
codes for determining the acceptability of stresses, and the need for
development of standard piping configurations appropriate to the
process. We will now develop these issues further into their practical
application.
Firstly, here is a checklist of the basic requirements that should be in
place before starting detailed piping design.
• P&I diagrams (certainly finalized to the point of including all but
a few small-bore lines) and the corresponding line list information.
• Piping design criteria, including a library of pipeline material specifi-
cations, and the minimum contents outlined in Chapter 14.
• Plant layout drawings. There is an overlap in the following text
between layout drawings and the detail work, because the layouts
have to anticipate some of the details now discussed.
• Drafting system (part of overall design criteria).
• Piping materials management system, including a fully developed
interface governing the catalogue of piping components, and the
system for take-off of parts and their roll-up and communication for
purchase and construction.
Nine basic observations on the design of individual pipelines follow.
• The pipes must fulfil the requirements of the process, in particular,
the slope, drainage, and venting arrangements must be acceptable.
The piping must be correctly configured for in-line instrumentation
and sampling devices.
• Most piperuns are horizontal, or at a small inclination to the
horizontal as required for drainage. By their nature, therefore,
pipelines tend to block off plant access, and putting the pipes in a
222 Handbook for Process Plant Project Engineers
perspective of the whole plant, which may be divided into process areas
(or units), in-between-process areas, and offsite areas. The most
common solution to running pipework between process areas is to
mount it on elevated piperacks, permitting access underneath. Offsite, it
is normally acceptable to run the pipes on sleepers or low supports at
ground level, and make appropriate arrangements (road bridges or
conduits) for crossing roads. Such arrangements may also be acceptable
in certain cases between process areas. The configuration in plan of the
piperacks and sleeper racks (to minimize their length), is a fundamental
layout issue which we will not revisit here; we will move on to addressing
the vertical planes.
Basic layout considerations will have established which pipes run on
which piperacks, and to these should be added an allowance for future
development and an understanding of what other services may run
along the piperacks, for example cableracks. The next step is to establish
the sectional arrangement of the piperacks and their contents, noting
the following piping groups which must be especially catered for:
1. pipes which must slope;
2. hot pipes, which require insulation and provision for thermal
expansion;
3. pipes which require access for cleaning or occasional dismantling,
including ducts which require entry for cleaning or refractory main-
tenance, and high-maintenance lines (such as slurry lines) requiring
occasional replacement;
4. plastic and small-bore pipes, which require special supporting
arrangements.
The first group is the most difficult. It is seldom a viable proposition
to slope the whole piperack. If there are only a few pipes that slope, it is
sometimes possible to run them on the outside of the piperack. The
more usual solution is to suspend them by hangers of appropriately
varying length. If the runs are too long it may be necessary to make
arrangements for one or two intermediate drainage points, and corres-
ponding reverse slopes or risers, to avoid the pipes from becoming too
low (or starting too high). In any event, these issues must be addressed
at the outset of piperack layout.
We will assume in the following that thermal expansion must be
accommodated by designing sufficiently flexible pipe configurations –
expansion joints are usually not permissible because of reliability
concerns, deposition of process materials in the pockets of the joints,
and the difficulty of catering for the end-thrusts created.
224 Handbook for Process Plant Project Engineers
All drawing offices involved in the design of hot piping need to have
tables of standard configurations of pipe loops and bends, and perhaps
a few more complex shapes. The tables should contain formulae or
graphs to determine the values and acceptability of the movements,
stresses, and anchor loads, depending on the temperatures, materials,
and pipe-sizes employed. These are all that are needed to design most
hot piperuns, dividing the pipelines up into sensible sections which give
attention to anchoring, guides, sliding supports or shoes, and adequate
end-clearances, all of which should be covered by drawing office
standards. Code requirements or common sense may dictate that
the designs should be subsequently analysed by formal flexibility
calculation and stress analysis (invariably now by the use of proprietary
computer programs), but except for the more intractable applications,
this exercise should be a confirmation of what has been adequately
designed in the drawing office.
The above process will yield a sensible system of configuring
the pipelines into bends and loops that are sufficiently flexible. The
ensuing preliminary pipeline designs must then be married to the
piperack layout.
Junctions between piperacks, and external connections to piperacks,
are facilitated by a change of elevation. Where there are a number of
pipes lying alongside on the rack, clearly only the outermost pipe can
receive a horizontal connection without obstructing the other piperuns.
It may pay to put a large commonly intersected line (such as a cooling
water header or common vent or flare line) on the outside for this
purpose, but in general, it is convenient to intersect lines vertically,
and therefore to intersect piperacks at different elevations, to prevent
obstruction (see Fig. 19.1).
In the case of pipe loops it is sometimes possible to arrange
all the hot pipes on one side of the rack, and to accept that all the
hot pipes will be looped at the same point along the piperack, in which
case the loop may be in the same plane as the rack pipes. Otherwise,
and if there is any doubt about possibly conflicting design develop-
ment or future needs, the loops should be stationed in a horizontal
plane above the rack pipes, and the pipes connected vertically to the
loops.
Moving on to the next two pipe groupings, the arrangement of access
when required for cleaning or dismantling should need no elaboration.
Plastic and small-bore pipes require an intermediate support system.
One design solution is the use of cable trays to support the pipes along
their length.
Detail Design and Drafting 225
19.3 Vessels
Unlike the connective items of steelwork and pipework, vessels are
essentially equipment items, and are therefore of less significance
in a text principally aimed at overall engineering rather than
discipline specialization. Vessels may well be purchased in the same
way as an equipment item, by the issue of a specification and a data
sheet giving the required dimensional outlines and process conditions.
This is particularly advantageous for pressure vessels, which are
invariably subject to a design code which embraces both design and
manufacture. These are interrelated, and it is not a good idea to split
the final design and manufacturing responsibility. By ‘final design’
is meant the detail design of a vessel for which the purchaser has
specified the following.
• The service conditions for which the vessel must be suitable. These
inevitably include ‘normal’ and ‘design’ figures, to allow for process
fluctuations. The ‘design’ pressure allows for the operation of safety
devices in a way that is prescribed by the code.
• The materials of construction and corrosion allowance. Normally the
purchaser (and ultimately the party responsible for process technology)
is expected to be more knowledgeable than the vessel vendor about how
to combat corrosion in the service environment, and takes responsibility
for this aspect. This can also be achieved by the specification of generic
types of materials which are acceptable (for example low-carbon steel),
leaving the vendor to decide within this envelope on the most cost-
effective solution for the service conditions stated, subject to the final
approval of the purchaser.
• The functional requirements of the vessel: capacity, general shape
and dimensions, internals, type and overall dimensions of support,
piping and instrumentation connections and access openings, lifting
lugs, and positions of pads and brackets for miscellaneous steelwork
connections.
• The code(s) of construction or list of acceptable codes.
228 Handbook for Process Plant Project Engineers
Support base design must also facilitate the initial positioning of the
supported equipment, which may be problematic if for instance cast-in
foundation bolts are employed. The designer should ensure that the
system of dimensional tolerancing matches the needs of the supported
equipment. For example, if two related machinery items have a
small specified tolerance on their relative positions, the feature of the
foundation design that limits the machinery positions, say the spacing
of the cast-in foundation bolts, should be directly dimensioned one to
another. If the relative positions are determined by dimensioning
each set of bolts from an individual machine centreline, the possible
cumulative tolerance error is tripled.
Where the dimensional tolerances of support bases appear to require
the use of special construction methods or templates, it does no harm to
call for that on the drawing. If the contractor comes up with a better
method it will surely be agreed on site, while if no special notes are
made the contractor may simply fail to meet tolerances which are
considered to be over-demanding.
A design practice is necessary for foundations subject to vibrating
loads. Usually the concern is simply to ensure that natural frequencies
of the supported system are well away from the operating frequencies.
Various design packages are available for such checks, but the inherent
variability of soil properties (initially, as well as over time) detracts
from the precision that can be offered by analysis. For many cases it
is possible to rely on design practices by which rocking motion will
be minimized, in particular to design foundations and bases which
have a low ratio of height to width in the plane of excitation. For more
critical applications, the involvement of the equipment vendor should be
secured as part of his contractual responsibility.
Transmission of vibration to other structures can be a problem. Even
on relatively simple applications, such as centrifugal pump bases, it
is necessary to minimize the transmission of vibrations (usually by
providing separate bases), otherwise standby machinery may be subject
to bearing damage. In the case of control rooms, any noticeable
transmitted vibrations are likely to be cause for complaint. Even when
environmental standards concerning permissible control room vibration
levels exist, and may seem attainable without special construction, it
is advisable where possible to be conservative. Positively isolate the
foundations and employ brick or concrete building methods if there
is any vibration source in the vicinity, especially for low-frequency
vibration exciters such as crushers and mills.
Most civil items within a process plant should be the subject of
Detail Design and Drafting 233
237
forethought and planning
238 Handbook for Process Plant Project Engineers
properties depending on their source, and may react with the process
materials. Special attention may be needed to follow up from the
design to the refractory installation, which is an art rather than a
science.
• Heating, ventilation, and air conditioning.
• The design of systems for fire prevention, detection, and extinguishing.
• Corrosion, and corresponding metallurgical and corrosion-
protection issues.
• Noise control.
• Hazard analysis and mitigation.
• Statutory compliance.
• Pressure vessels – possibly a code requirement.
In general these needs should become apparent from an appraisal of
the critical aspects of plant design, and of the resources available.
Chapter 20
1
Predictable, with experience. The number of internally generated documents should
include a growth provision (mainly for extra small purchase orders and contracts), but
the major underestimation is usually in vendor data (drawings, manuals, etc.); if in
doubt, refer to similar previous work to estimate these.
Traditional Documentation Control 245
Fifth Cycle
More on Management
This page intentionally left blank
The Organization of Work 249
Chapter 21
hours of the day or weekend into which to expand the work. Paying
for work by the month, and then paying overtime or other benefits
for overtime, is the same as paying by the hour.
• Performance of all activities is enhanced by competition, provided
that workscope is clearly defined and product quality is adequately
maintained by setting acceptability criteria and by inspection.
• Working efficiency is improved, and errors are reduced, by adopting
standardized designs and working procedures.
These observations are surely what common sense would predict, and
yet in many instances plant owners and project managers continue to
believe that value may be obtained otherwise. For instance, they may
blindly put their faith in professional people paid by the hour, and be
surprised when the work difficulty increases and it has to be performed
at a rate of 70 h per week or more. Or lump-sum contracts may be
awarded on a highly competitive basis, without in-depth surveillance
and inspection of the work, with the result that all the gains of low
price are lost in poor performance and poor quality. The delivery of
inherently poor quality, even if not discovered until after payment has
been made, may seem to be a simple problem to rectify legally, but
in practice it is usually accompanied by incessant argument and
obfuscation. Quite often, a poor-quality supplier or contractor concen-
trates his attention on surviving in the relationship long enough until
any move to transfer the work to a competitor would be intolerably
disruptive to the project. He then utilizes this position to negotiate
better terms, failing which work performance will further deteriorate
or stop. Here is a dilemma for the project manager: how does he get
the best prices for his DFC components, without having to spend,
what was thereby saved on increased surveillance, inspection, and
contractor substitution costs, and increasing the IFC cost component?
He must find the best balance, which will not be the same in different
circumstances.
In the light of the above, let us now examine some of the issues which
may arise from the way in which the performance of work is organized
for major projects. The classic system of organization is based on the
employment of an overall ‘owner’s engineer’, which/who may be part of
the owner’s organization, or a consultant or possibly an engineering
company. This is the contractual engineer who is charged with:
• the conceptual design and control estimate;
• the definition of packages of work, which together fulfil the project
objectives (or of a single turnkey package);
The Organization of Work 251
1
‘Consultant’ in this context may be an in-house role: whoever has the technical ability
to, and is charged with, appointing one or more contractors to carry out the project
work, or indeed, setting up an in-house team for the purpose.
252 Handbook for Process Plant Project Engineers
• Motivation:
– Are the conditions of employment directed to motivate the
work-groups to meet all the project objectives?
– Are there any additional low-cost benefits which may signifi-
cantly promote productivity and quality?
• Communication:
– Are important interfaces connected by direct communication?
• Standardization:
– Are work practices, documentation and numbering systems, and
design features standardized as much as possible across the
work-groups?
• Duplication and omission:
– Are any functions duplicated?
– Are any requirements of the overall project objectives and
workscope omitted?
• Organizational hygiene:
– Is the proposed overall administration system likely to cause any
unnecessary frustrations?
• Management complexity:
– Has there been a detailed evaluation of the full extent of overall
management work and manhours to adequately manage the
proposed work structure?
– Is the cost and effort justified by the benefits of the proposed
organization?
21.2 Procedures
‘Procedures’ is a surprisingly controversial subject, in that one
encounters quite a few apparently competent people who are quite
antagonistic to any form of written procedure, and use words like
‘red tape’ to describe them. And yet a procedure is just a way of doing
something, no more and no less. As there are inevitably good, bad,
and unacceptable (even dangerous) ways to perform project activities,
developing and using sound procedures is an inevitable necessity.
Furthermore, to work in a team where efficient communication and
co-operation are essential, and where work is planned and controlled,
procedures need to be uniform.
Undoubtedly the reason for widespread apathy and occasional
antagonism on this subject is that, so often, procedures contain
unnecessary verbiage and unnecessary activities, and generally over-
complicate the work. And if this is the reason why work is often poorly
The Organization of Work 255
Chapter 22
Construction
1
Factors such as environmental impact are considered here to be the responsibility of
another specialist consultant.
Construction 259
2
This subject is discussed in Chapter 25, Communication.
260 Handbook for Process Plant Project Engineers
3
It is possible for the project engineers to order long-lead bulk items (particularly
valves, which can be ordered early from P&ID information), and hand them over to the
construction contractor, who must procure the balance of items. This is indeed sometimes
practised, but it brings problems of its own, in particular a duplication of systems.
262 Handbook for Process Plant Project Engineers
4
Over the ages many consulting engineers, in particular, have made an art form of such
contract clauses, tested in court following legal dispute and found to be effective.
Construction Contracts 265
Chapter 23
Construction Contracts
1
This and similar observations have often led the author to wonder whether the whole
process of bidding for lump-sump or rates-based work of this type is not really a
complete farce. Perhaps both parties know (but will never admit) that there are so many
intangible and unquantifiable factors affecting work performance that the end price will
be fixed by negotiation, on a basis of what is ‘reasonable’. If so, the observations made
in the previous discussion on relationships are reinforced.
Construction Contracts 267
2
Or a multiplier based on standard ratios of artisan’s time per operation: this simplifies
the number of different rates to be determined, but not the surveyor’s work in determining
the quantities.
268 Handbook for Process Plant Project Engineers
3
The moral of this is that separate payments should not be made for any activities,
tools, or incidentals which are not a part of the end product: pay for the end product,
not the means of getting there. It may be necessary to ensure by specification and
by supervision that no unacceptable short-cuts are taken as a result, for example, that
scaffolding is adequate for safety. If scaffolding is needed in the same place by more
than two contractors, let them strike a deal between themselves.
Construction Contracts 269
4
At least at the project management level. The individual contractor will still require,
for his own management purposes, documents such as quantity schedules for ordering
materials and controlling work. He will have to generate these documents, which would
otherwise be prepared by the engineer. Failure to recognize this can cause both dispute
and substandard work performance due to working without proper controls.
270 Handbook for Process Plant Project Engineers
23.3 Claims
The settlement of contractors’ claims can be a nightmare which turns
an apparently successfully completed project into a disaster. Claims
settlement is one of the most frequent causes of major cost overrun.
Even if the claims are disputed and eventually legally rejected, the
length of time taken and the effort involved can leave, over an otherwise
successful project, a cloud of uncertainty that remains until long after
the project achievements have been forgotten.
There is not much defence against the institution of unreasonable
claims, other than the avoidance of litigious contractors. But such
claims are likely to be seen for what they are, and are unlikely to be the
subject of reservations as to the project outcome.
Evidently the subject of claims management is an important one,
worthy of specialized input which is beyond our present scope. For most
purposes, project engineers concentrate their attention on rapidly and
reasonably settling any problems which may otherwise escalate into a
contractual dispute, and above all on eliminating the causes of such
problems.
Construction Contracts 271
The main causes over which the engineer can exercise some control
are the following.
• Unclear and possibly inadequate work description (for example,
of any of the elements listed above). Work should be described
inclusively, that is to say that all aspects not specifically excluded are
included. Let the bidders qualify their offers if they are in doubt.
• Late provision to the contractor of information, materials, facilities,
or access (including access to work which first has to be completed
by others).
• Changes or faults in information and materials supplied.
We will not dwell further on the subject of work description, but
rather concentrate on the factors of lateness and error. These are in
practice quite closely linked: working under pressure of time results
in increased error. Obviously, one way to minimize potential claims
from this source is to delay construction work until engineering work is
complete and all materials and equipment are available. This generally
is not an acceptable alternative, because of both the extra cost of
financing an extended project, and delays in product availability. On the
contrary, it is standard practice to appoint a construction manager very
early in the project, even at the outset. This person will want to get on
with the job, and be almost guaranteed to create the maximum pressure
to get contractors on site and commence construction work.
If a project deteriorates to the point where both the quality and the
timing of information and materials sent to site are inadequate, it can
and usually does deteriorate in exponential style as more resources are
diverted to the solution of problems and less are available for the
advancement of work which is already behind schedule. In this context,
construction claims and costs are bound to soar owing to the combined
effects of disruption and extended site establishment.
The consequences of allowing the provision of materials and
information to site to get out of hand in this fashion are usually far
more severe than the consequences of initiating a controlled schedule
delay. Unfortunately this is not readily perceived early enough, possibly
out of optimism, more probably from collective behaviour in a climate
where the acknowledgement of bad news is regarded as defeatism.
There can be many causes of unforeseen schedule delays, ranging
from lack of competence of the engineering and management team to
supply and shipping problems and unrealistic schedules in the first
place. Another frequent cause is the acceptance of too many changes in
plan, especially of design changes after detailed design has commenced.
272 Handbook for Process Plant Project Engineers
Chapter 24
Commissioning
the task, and the plant design engineers are given this responsibility,
which has a logical overlap. Some organizations employ specialized
commissioning engineers, at least to lead the commissioning effort.
This can be a very good investment, especially for a large plant where
by providing for dedicated responsibility and focus, significant improve-
ment on schedule and avoidance of negative start-up incidents may be
achieved. A day extra taken up during commissioning is the same to the
plant owner as a day extra taken up during design or construction; in
fact it is likely to cost more, as the plant owner’s commitments in terms
of product marketing and operational costs are likely to be higher.
Because commissioning comes at the back end of the project there is a
danger that the work may be under-resourced, because the funds have
been pilfered to pay for budget overruns. It is essential to comprehend
the scope and length of commissioning activities and include them in
the initial project plan and budget allocations, and ensure that this
commitment is maintained.
Detailed planning of commissioning and plant handover is as essen-
tial a component of the overall project plan and schedule as any other
grouping of activities. Inevitably, some part of the critical path goes
through commissioning up to handover. Like all planning, it is reliant
on determining the methodology and procedures to be utilized, as well
as the work breakdown, the activity durations, and the sequence logic
which arise. These issues must be developed and agreed before it is
possible to have confidence in the schedule leading up to plant operation.
When drawing up detailed commissioning plans and checklists, there
are two important sources apart from previous experience:1 firstly the
process technology, and secondly the equipment vendors. With regard
to the latter, it is necessary to review in detail the instructions provided
in the operating and maintenance manuals, which must therefore be
available well before the time of commissioning. The actual work of
reviewing the manuals, clarifying any problems, and drawing up appro-
priate plans and checklists can consume many hours, but it must be
done properly. Especially for high-cost plant items, it is often considered
to be worthwhile to bring vendor commissioning engineers to site,
thereby introducing an experienced specialist and reducing the load on
the project commissioning team. Responsibility should be given to the
vendor for ensuring that the equipment has been properly installed, thus
1
Previous experience may be supplemented by industry literature such as the relevant
API standards.
Commissioning 275
Chapter 25
Communication
response to the needs of the team. They may be wasted unless there
is a dedicated responsibility to ensure that documents go back and
forth as ordained.
• Control of external correspondence, and specific procedures for
specific types of communication. All external communication,
particularly communication with the client, vendors, and contractors,
needs to be the subject of procedures controlling formulation,
authorization, record, and (where required), proof of transmission
and prompt for follow-up.
• Meetings, generally scheduled weekly and monthly: a heavy but
necessary drain on management time.
• Reports, concerning in particular: meetings; activities such as
inspection, expediting and site visits; project controls such as cost
reports; and project progress reports to the client.
• Documentation indexes and project component registers. The
indexes enable users to know what documentation exists, and to
access it. Registers are indexes to information which is not the
subject of a specific document (such as equipment lists), but the
nomenclature is not standardized: it is quite common to use
‘register’ and ‘index’ interchangeably. Equipment registers or lists,
giving a summary of the main characteristics such as mass or power
rating, are essential control documents referenced by all disciplines.
• Briefs, discussed below.
A brief is essentially an aid to planning an activity, and as such it is
part of the basic project execution philosophy of ‘Plan the work then
work the plan’. Its objective is to ensure that all important aspects of a
job are understood and communicated before the job is performed.
The effectiveness of briefs is enhanced by the use of checklists, which
serve not only to remind the author of the brief’s essential contents, but
also to prompt him to ensure that he has himself completed all the nec-
essary preceding activities. Here is a general form of checklist.
• Description of activities to which the brief refers.
• List of reference documentation (relevant information and data,
work instructions and applicable standards, and acceptability cri-
teria) needed for the task.
• Any special materials, equipment, skills, or methods to be employed
when carrying out the activity.
• Any special hazards or safety precautions.
• Note of critical aspects of the activity.
• Note on interaction with other activities.
Communication 279
Chapter 26
26.1 Change
We started our review of project management, in Chapter 3, by stressing
the fundamental need of planning and working to plan. Inevitably there
will be a need to change the plan. It is unlikely that every event will
materialize as foreseen, or that arising from the unforeseen, there will be
no previous decisions or designs that subsequently are seen to be wrong
or sub-optimal.
As we have seen, there are indeed bound to be many assumptions
made during planning that will inevitably lead to some measure of
revision and re-work. The issue is not therefore whether to permit any
changes, but how to control them so that they do not destroy the
manageability of the project.
Change control has a few components: change recognition, change
evaluation, change approval or rejection, and change implementation,
including the consequent revision to plan. These are discussed below.
• Change recognition may sound an odd subject to the uninitiated,
but it is a fact that changes are too often made inadvertently or
surreptitiously, with damaging consequences. An error is in general
a form of inadvertent change from what was intended. Once an error
has been made, it may be discovered and consequently the relevant
work may be surreptitiously changed, with unforeseen repercussions.
Process plant design and construction is such a heavily interlinked
exercise of different activities and disciplines, that the consequences
of change are very easy to underestimate. In fact ‘change’ is
invariably one of the most emotive issues in the business.
Change recognition is mainly dependent on the training and
conscientiousness of the work performers and their supervisors. It is
282 Handbook for Process Plant Project Engineers
26.2 Chaos
The following message may be found displayed at the workstations of
many process plant project engineers:
‘From out of the chaos, a small voice spoke to me and said Relax and
be happy, things could be worse. So I relaxed and was happy … and
things got worse.’
284 Handbook for Process Plant Project Engineers
1
Work breakdown structure, as planned.
Change and Chaos 287
Chapter 27
Fast-Track Projects
than the piping being designed to suit the steelwork after its structural
design, as in Fig. 4.1. The efficiency of this process is dependent on such
factors as the quality of teamwork, the design aids available to assist
co-ordination, and the success with which the layout drawings have
anticipated the final structural design.
The practice of parallel working extends into the manufacturing and
construction phases. Examples are:
• equipment design and manufacture commencing, while secondary
equipment design information, such as positions of piping connections,
is still ‘on hold’;
• placing of contracts for construction work before the drawings to
define the details of the work have been completed; and
• the construction of earthworks proceeding before the plant layout
details have been finalized.
In the making of assumptions of design information, usually related
to information from a third party, the progress of design work is based
on the best (or sometimes most conservative) estimate of the information.
For example, structural steelwork design may be based on the anticipated
mass and dimensions of a supported equipment item for which final
vendor’s information is not yet available. Evidently the accuracy of
the assumptions, the designed-in flexibility to accommodate incorrect
assumptions, and the ability to influence the equipment item vendor to
comply with the assumptions (or the freedom to choose a compliant
vendor) will affect the efficiency of this process.
If the assumptions to be made relate to the performance of work by
another project engineering discipline, say if the mechanical engineer
orders a pump before the process engineer finalizes the calculations
which will affect the required performance, it is clearly only a matter of
semantics whether the practice is described as an assumption or parallel
work.
If work proceeds in parallel or by assumption, there is clearly a
possibility (for a large amount of such work, a certainty) that there will
be some errors and inconsistency. Time and resources must be allowed
to detect the problems and correct the work in the design office, redesign
the equipment being manufactured, modify the construction on site, or
a mixture of all three.
The practices outlined above are effectively a means of breaking the
logic of schedule dependency; there are also other ways of going about
it. One method quite often practised at the civil/structural-steel interface
is to work ‘from both ends towards the middle’. For instance, there may
Fast-Track Projects 291
problems and opportunities which can arise, here are the outlines of two
projects where a fast-track approach was adopted.
agreed with him on an absolute maximum 10-day period for the assess-
ment of equipment bids, and 1 week further for negotiation and order
placement, and also suggested that a few traditional bidders be eliminated
because of a record of tardy performance.
FF got their plant commissioned within the 12-month period – but
not without additional, unforeseen costs and problems. Construction
plans had to be changed a few times, as insufficiently thought-out
designs had to be modified in the field, but it was no coincidence that
the errors occurred in the less critical areas where there was some
flexibility to make changes. And undoubtedly, some premium was paid
for abbreviating the procurement deliberations and for excluding
vendors and contractors who were adjudged unable to meet the schedule
(although most of them claimed otherwise). The extra costs of fast-
tracking the project were considered to be far less than the commercial
benefits of early product delivery.
Chapter 28
Final Cycle
Strategies for Success
This page intentionally left blank
Project Strategy Development 305
Chapter 29
Project Strategy Development
It may seem odd that this subject, which is the essence of deciding how
to go about the whole project, should be left to the end. (The final chap-
ter which follows is essentially a summary.) The reason is that it cannot
be addressed economically until all the main issues have been exposed
and discussed.
The project strategy is the name given to the highest level of a project
plan. This word tends to be rather misused or abused in management
literature. What it has in common with the military usage, from which it
is derived, is that it addresses an entire plan of campaign (the project).
Any specific issues, such as construction or engineering, should only be
addressed in so far as they directly relate to the overall objectives and
cannot be settled in a smaller forum. The introduction of too much
detail, or too low a level of planning, inevitably detracts from focus on
the major issues.
For instance, consider the means of addressing design criteria (the
highest level of engineering planning) during project strategy develop-
ment. With the exception of the most fundamental issue, the overall
plant performance definition, the design criteria should be addressed
as to how they will be formulated (which will be discussed below)
rather than their content. There may well be lower-level strategies than
the project strategy. For instance, it may well pay to organize a multi-
functional strategic meeting to discuss the way of formulating and
negotiating a particular construction contract within the overall project.
The usual objectives are to ensure that the important engineering,
procurement, and construction implications are understood and to derive
an appropriate overall strategy, before starting any one function or
discipline’s work. However, such supplementary strategizing is conducted
with the overall project strategy as a base; it is probably a misuse of the
word strategic, but commonly employed when one wants people to
‘think big’. ‘Big’ is of course always relative.
306 Handbook for Process Plant Project Engineers
course of the project, and very damaging they can be at a later date.
If the answer to the second part of the question is positive, subsequent
action is fairly obvious:
• evaluate the opportunities fully;
• verify that the improvements are real;
• decide how their introduction can best be utilized for ‘our’ benefit;
• negotiate the changes.
In the case of the project team undertaking work in accordance with
client specifications, obviously the probing and full understanding of
the specifications/contract is an essential prerequisite to this part of the
strategic planning process.
Chapter 30
Appendix 1
Jargon
Appendix 2
The same care is required as in the case of the product. In the case of
purchased utilities or reagents there may be leeway for the project
engineers to determine the specifications, but there will inevitably
be limitations (for example of price or source) which should be
ascertained at the outset.
• Limitations on plant effluent, waste disposal, and environmental
impact.
• The plant capacity (its rate of working). This may be defined
in terms of quantity of product, quantities of types of product,
quantity of feedstock, or a combination of quantities to be produced
and processed over a period of time.
‘Period of time’ needs to be qualified, to take into account: the
plant operational and maintenance requirements; whether operation
is continuous or how many hours per week or year; and the time
allowed for scheduled and unscheduled shutdowns for maintenance,
inspection, and repair (collectively referred to as ‘maintenance’). The
starting point for plant design is usually the ‘average’ capacity (over
a long period of time which includes all shutdowns for whatever
reason). A ‘normal’ capacity is then calculated for the time when
the plant is in operation, allowing for shutdowns according to the
defined plant operational mode and the time to be allowed for
maintenance; these have a major bearing on the plant design and
reliability requirement.
• The required flexibility of operation. The range of capacities
over which the plant must be operable (or the ‘turndown’) is the
most important consideration, but all required abnormal opera-
tion should be considered, including initial start-up, normal
start-up, normal shutdown, emergency shutdown, and power
failure. Sometimes a ‘maximum’ capacity may be defined to pro-
vide an extra margin for operational contingencies; however,
care must be exercised in the use of this term, as previously
outlined.
• Plant reliability, maintainability, and life requirements, consistent
with the plant capacity calculation.
• Requirements for the operation of the plant, limitations on
the numbers and skills of operational personnel, degree of
automation, and local or statutory regulations which may affect
plant operation.
• Special safety requirements, hazard containment, and fire prevention
and extinguishing.
• Plant performance testing and acceptance standards.
Design Criteria Checklist 319
Index
Abrasion 186, 187 safety during 276
Access 112 Communication 34, 277
Accuracy 94, 97, 99, 100 Competitive bidding 61, 67
Agitators 189–191 Compressors 164–167, 171
Air slides 179 air 163, 164
American Petroleum Institute (API) 166 centrifugal 164
Approval 77, 143 gas 171
ASME Standard B31.3 167 turbo-compressors 162, 165
Assumption 38, 289–291, 312 Computer systems 300
Audit trail for design information 43 Computer-assisted design 113
Authority levels 256 Conceptual designs 25
Availability 16 Conceptual study diagram 39
Average capacity 318 Construction 28, 33, 257, 259
contract specifications 265, 266
Batch production 16 management 257
Belt feeder 183 Consultants 238, 241, 312
Bins 181–183, 229, 230 geophysical 238
Block plan diagrams 108, 109 bulk solids flow 180
Brief 278, 279 Consulting engineers 251
Bulk solids transport 107, 173, 175 Contingency 101–103, 311
angle of repose 174 Contractors 59, 70, 72, 265
flow 173, 182 Contracts 59, 61, 63, 65, 68, 265, 269
Control 27, 212, 214
Cash flow 83 system design 14, 233, 235
Change 25, 46, 75, 271, 272, Conveyor Equipment Manufacturers’
281, 283, 285, 286 Association of USA (CEMA)
approval/rejection 282 175, 178, 180
control 281 Conveyors 175, 184
evaluation 282 belt 178, 179, 184, 206
implementation 283 chain 179
recognition 281 pipe-belt conveyor 179
Chaos 25, 281, 283, 284, 287, 288, 310, 313 screw 180
Chemical plant 53 troughed belt 175
Chezy formula 196 vibrating 179
Chutes 174, 176, 177, 183, 229, 230 Corruption 60
blockage 174, 175 Cost 41
Civil design 230 capital cost 77, 78, 144, 145
Civil works 208, 209 control 33, 41
Claims 270–272 direct costs 57, 58
Client 51, 55, 65, 72 direct field cost (DFC) 59
Codes of practice 131 engineering cost 84, 86
Commercial appraisal 144 estimation 56, 94–97
Commercial relationship 60 indirect costs 56–58
Commericial policies and procedures 61 indirect field cost (IFC) 84
Commissioning 29, 273–276, 280 maintenance cost 82
cold 274 management 41
hot 274 manipulation 270
plans and checklists 274, 277 operating cost 77, 78, 143, 144
324 Index
Safety 16, 123, 125, 127–132, 183, 184 Valley angle 182
Schedule 41 Value engineering 117–119, 139
shortening 38, 289, 291 Vane angle 155
Screw feeder 183 Vendor 142, 143, 145, 150, 274
Semi-continuous operation 16 documents 137
Sequence planning or scheduling 35 drawings 46
Setout 258 Vessels 226–229
Sewers 201 Vibrating loads 232
Silos 175, 181
Single line diagram 211 Wear at the flow boundary 174
Site 319 Work organization 27, 249, 250, 252
office relationship 262 Workflow 255
conditions 140 Workshop detailing 218
This page intentionally left blank